Hot 100 Weekly Top 10 Points Estimated

[fhas]

So, I've decided to go back to good old days and do the estimated points and the component breakdown with the sales-radio-streaming ratio for the top 10 songs of the week. I used to post the numbers on the Year-End Thread, but I think it's better to have a thread only for the Top 10.

The sales numbers aren't as accurate as it used to be because BuzzAngle's sales data is not matching Nielsen's anymore.

Streaming numbers are more accurate than before because certain forms of UGC (user-generated content on YouTube) were removed and the other forms of UGC were downweighted.

This is more accurate than a typical prediction because I adjust the numbers based on Billboard's Digital Sales, Radio Songs and Streaming Songs charts as well as Billboard's Chart Beat articles and a few other sources.

SALES: 10% margin of error

RADIO: 3% margin of error

STREAMING: 10% margin of error

HOT 100 POINTS: 5% margin of error

Beginning with the chart week of January 25...

[fhas]

[fhas]

[fhas]

[fhas]

[fhas]

[fhas]

[mikerivera]

Really glad you’re doing this. I notice that no one is currently doing real points projection of the 2020 charting year, so this will help massively in anyone who wants to start doing it.

[fhas]

Feb 16, 2020 14:48:20 GMT -5 mikerivera said:

Really glad you’re doing this. I notice that no one is currently doing real points projection of the 2020 charting year, so this will help massively in anyone who wants to start doing it.

Thank you!

Last week, there was good example of how real points are interesting: everything i wanted (#10 - 20,000 points) was closer to Hot Girl Bummer (#19 - 16,700) than Don't Start Now (#9 - 23,700) in chart points. Also, EIW was 3,900 points weaker than the #10 song from three weeks ago (Yummy - 23,900).

[Baby Yoda Hot100Fan]

^Agreed. Weekly positions in any chart give us an idea of how whatever's charting did that week in relation but doesn't actually tell us how close it is to other positions. Life is Good could have debuted at #1 had it been released during one of those weeks for which the #1 was pretty weak in points. So peaks can be deceiving as well.

[fhas]

[fhas]

[fhas]

^I've been doing this for more than 100 weeks (July 2017-August 2019 and January 2020-present) and this is the biggest week for radio in the top 10 with 39.5% of the points. The previous biggest week was on January 19, 2019 (Sunflower's only week at #1) when radio's percentage was 38.3. That week was High Hopes' peak in audience impressions with 131.7M.

[fhas]

[fhas]

[fhas]

[fhas]

[chanman]

I know actual points not rank, is the more important one. Since you have shown the Ranks....can it be viewed that a pre-requisite to be in the top 10 is to have Top 50 (Airplay) and Top 25 (Streaming)

[fhas]

Mar 1, 2020 13:06:16 GMT -5 chanman said:

I know actual points not rank, is the more important one. Since you have shown the Ranks....can it be viewed that a pre-requisite to be in the top 10 is to have Top 50 (Airplay) and Top 25 (Streaming)

It's a good question, but I don't think it's a pre-requisite.

If the song is huge on streaming or if it's very strong and has good sales numbers, airplay is not needed for top 10.

Godzilla debuted at #3 with 41.1M streams + 24k sales and only ~1M audience impressions.

Life Is Good debuted at #2 with 50.7M streams + 25k sales and ~15M audience impressions.

Also, Memories was #23 on Streaming Songs last week, but if it was #50 (or just outside the top 50) it would still be ranked in the top 10 on the Hot 100.

I never realized that a song doesn't need to rank in the top 50 on streaming to reach top 10 on the Hot 100. In 2020, that's kinda absurd and I'm scared lol.

[chanman]

I wasn't factoring in Sales Rank because it's contribution is usually around 10% or lower. But then, I am glad for the example, you have shown with Godzilla (even though I don't like the guy, hahaha). Then again, sales wasn't needed according to your formula ...Godzilla would give 24,000/5.4 = 4444.444. Then 30,800 - 4444 = 26,356...…...….a number #5 debut instead.

Well look at that....streaming alone could probably be enough for a top 10. I guess it also depends on how the other artists songs are doing in their categories to compete. I would say Eminem was lucky enough that week to get a high debut on practically streaming only. Good job in presenting your data. That's how we understand data better and better to predict top 10 for the next week.

It seemed that both Life is Good and Godzilla on the strength of streaming alone would have still made top 10.
My next question to analyze further... what would the total streaming numbers need to be, to at least reach top 10 without airplay or sales, on average? 40 million?

[fhas]

Mar 1, 2020 14:21:40 GMT -5 chanman said:

I wasn't factoring in Sales Rank because it's contribution is usually around 10% or lower. But then, I am glad for the example, you have shown with Godzilla (even though I don't like the guy, hahaha). Then again, sales wasn't needed according to your formula ...Godzilla would give 24,000/5.4 = 4444.444. Then 30,800 - 4444 = 26,356...…...….a number #5 debut instead.

Well look at that....streaming alone could probably be enough for a top 10. I guess it also depends on how the other artists songs are doing in their categories to compete. I would say Eminem was lucky enough that week to get a high debut on practically streaming only. Good job in presenting your data. That's how we understand data better and better to predict top 10 for the next week.

It seemed that both Life is Good and Godzilla on the strength of streaming alone would have still made top 10.
My next question to analyze further... what would the total streaming numbers need to be, to at least reach top 10 without airplay or sales, on average? 40 million?

Average is 35 million, but it changes depending if it's mostly paid streams or mostly free streams and if the top 10 is strong or week in points.

[fhas]

[fhas]

[fhas]

ON is top 10 with 56.8% of its points coming from sales.

The last time that a song was top 10 with more than 60.9% of its points coming from sales? Shallow's first week in the top 10 - Oct. 20, 2018 - 65.5%.

[kingvavis]

It's crazy how only four songs in the top 10 have more than 20 million streams. This time last year, every song in the top 10 reached those streams.

[fhas]

Mar 8, 2020 11:00:53 GMT -5 kingvavis said:

It's crazy how only four songs in the top 10 have more than 20 million streams. This time last year, every song in the top 10 reached those streams.

UGC rule change. Last year this week (March 9), the weakest song in the top 10 had ~25M streams (Happier). That's why I think they should've adjusted streaming's weighting when they changed the UGC rule.

[fhas]

[fhas]

[fhas]

Formula adjustment:

Well... I had to adjust the formula that I use for the Hot 100.

From: SALES/5.4 -- To: SALES/5

Reason: March 7 chart

BTS' On was #4 with 86k sales, 500k airplay and 18.3M streams (10.7M on-demand)
Dua Lipa's Don't Start Now was #5 with 10.6k sales, 90.7M airplay and 17.5M streams*** (13.8M on-demand)

***Don't Start Now actually beat On's streaming because it had way more paid streams than BTS' song.

Why it took me almost two weeks to adjust?
- I thought that I was overestimating DSN's numbers, so I updated its numbers down, but when Billboard posted the official numbers for DSN on the March 14 chart, I realized that I didn't need to adjust Dua's numbers and that what I actually needed to adjust was the formula to give more points to sales.

Also, I have changed the numbers for every week since January 25 with this new formula and the only other difference was two weeks when 10,000 Hours fell below Dance Monkey by less than 500 points, so I just adjusted the airplay numbers for both songs and it was fixed.

What does it mean in terms of points distribution?
Sales gain 0.5%, streaming loses 0.3% and radio loses 0.2% of the points.

[fhas]