(BIG UPDATE) 2010S Decade-End Hot 100 USING REAL POINTS

[Deleted]

Weighed peak points is another matter. It's the structural problem all over again. It's hard to say what for example Party Rock Anthem's peak point was since it got like 250,000 sales maximum a week not counting xmas week and 150-ish m audience impression at the outside. If you *1.55 it's something 60,000 points but I don't think I would feel nearly at much meaning it barely scraped the median what a no.1 song back then would look like.

I don't think I have the incredibly accurate numbers on exact weeks. If you're into it, you could do the calculations on peak weeks yourselves. Like I said, weighing is actually the biggest obstacle to put eras together

[inverse]

Just wondering where do you have If I Die Young? It should be kinda close right

[Deleted]

May 26, 2017 1:01:26 GMT -5 inverse said:

Just wondering where do you have If I Die Young? It should be kinda close right

It's about 710,000 point-wise. But I didn't really calculate much what's below top 100. It's not a small undertaking after all~

[Deleted]

It's not that hard. It's just the points before 2014 (especially before 2012) could more likely use rougher estimation than the years after since we got week to week prediction now. Can you find anywhere year-end predictions for the year of 2010, 2011, 2012, 2013?

[Deleted]

I've been also hitting upon 2000s decade-end hot 100 using real points. But that's not gonna be easy since I have to be messing around what Billboard been up to all those physical sales era / airplay / digital formula change and for example strange 2007 year-end chart and all. I don't know how to get more accuracy since especially 2000-2004 the single sales data are very hard to find and it's mostly airplay. But maybe I'll post something not just this decade

[inverse]

Also maybe I'm reading the numbers wrong but it looks like you have See You Again at 150000 in 2016 and Sail at 175000 in 2014. If those numbers are right then that seems like an error. I know that the cutoff point for the year end hot 100 varies from year to year but there's no way that the point total to miss the list in 2014 was a whole 25000, easily equivalent to a week in the top 10, less than the number 99 in 2016. In fact that's about equal to what Watch Me did in 2016

[Deleted]

If this is accurate then why does billboard say PRA is the 5th biggest song of all time? Why would they use an inverse ranking system instead of actual points? Does this also apply for year-end charts?

[Deleted]

May 26, 2017 15:33:16 GMT -5 inverse said:

Also maybe I'm reading the numbers wrong but it looks like you have See You Again at 150000 in 2016 and Sail at 175000 in 2014. If those numbers are right then that seems like an error. I know that the cutoff point for the year end hot 100 varies from year to year but there's no way that the point total to miss the list in 2014 was a whole 25000, easily equivalent to a week in the top 10, less than the number 99 in 2016. In fact that's about equal to what Watch Me did in 2016

I disagree. 
Look at 2016, the year end #100 was less than 160,000 points
But this year, the cut-off line is projected to be above 185,000 even with a 0.95X multiplier.

Increasing or adjustments of streaming and formulas can impose a big change to the bottom of a year-end list. That's why I weighed them by case. In 2011 the year-end #100 song was above 200,000 points after the first general weighing. But that's why I also fine-tuned them.

[Deleted]

May 26, 2017 16:10:53 GMT -5 @alec935 said:

If this is accurate then why does billboard say PRA is the 5th biggest song of all time? Why would they use an inverse ranking system instead of actual points? Does this also apply for year-end charts?

That's Billboard's choice. And not a good one imo. 
I've been a big fan of billboard's decade/all time charts for while but then turned to the real-point version since like I explained, the inverse point system gives unnecessary weights to top 10/no.1 weeks alone regardless of quality. Not to mention Billboard's sense of weighing and balance is... just incredible not in the right way

[Deleted]

PRA's more like around the #11-#16 biggest hot 100 song of all time invited comparison to #5.
You really should read more what I typed at the beginning of the thread/update before questioning.

[fhas]

I think positions are very important. It would be unfair if Cheap Thrills received the same points as #1 that Mask Off received few weeks ago as #6 and that's what would happen with real points. However, it's unfair that Hello's first four weeks at #1 receives the same points as Cheap Thrills' four weeks at #1, since Hello's weeks were much stronger. Solution: an inverse point system with bonus points as a percentage of the song's real points.



Example:

Weak #1

Cheap Thrills' 4th week at #1: 52,000 points (65,000*0,8) received based on an inverse point system + 8,950 points (25% of its weighed real points)

Total points: 60,950



Strong #1

Closer's 4th week at #1: 52,000 points (65,000*0,8) received based on an inverse point system + 17,600 (25% of its weighed real points)

Total points: 69,600

[Deleted]

So... your point is to pull up a balanced point system between real points and inverse point system? I thought about that, but oh geez that's gonna be complicated! How to balance? What to balance? How to weigh the eras then? ......
Not a good idea imo. Why sticking to the inverse point system when you can spend some effort to carry off the real points version like the year-end charts?
It's like saying, why sticking to a flawed simulation of truth when you can have truth itself? Of course the concept of "truth" or "truthful simulation or measurement" of songs' performances and success are complicated and almost implausible. But bringing an inverse point system in on this is only like, adding insult to injury. You know "most successful songs of all-time" is more like, smack-dab, how many copies are sold, how many times it's streamed or whatever and you go balance the data with other eras and that's it. If in the same era a song's sold 5 million copies but peaked at no. 8 vs. a song sold 1 million copies but peaked at no. 1 for 4 weeks, how can you look at it? Infinite amount of ways. But people usually just look at the peak and the final numbers on how well it's sold and that's it (but focusing on how well ultimately it's sold). Fixating on the simple quantification of top 10 week won't cut it. If it does, then, you can fixate on weeks spent in top 2, top 3, top 5, top 20, top 40, top 50, top 100, etc. What do you do? Run a survey on which one people give more crap about? Probably not.

[Deleted]

Did mask off actually have as many points as cheap thrills did when it hit number one? I'm pretty sure it was like over 5k (prolly 10k) less but I might be wrong

[nick64]

May 26, 2017 18:52:43 GMT -5 fhas said:

I think positions are very important. It would be unfair if Cheap Thrills received the same points as #1 that Mask Off received few weeks ago as #6 and that's what would happen with real points.

Why would that be unfair? If that's the amount of points a song earned, that's how much it earned. "Cheap Thrills" was lucky enough to have nothing do better than it that week, while "Mask Off" happened to have five songs do better than it. Doesn't change the fact that they both did about the same amount. It would be unfair to "Mask Off" to discredit it by making it worth less just because it ranked at #6 instead of #1.

[fhas]

May 27, 2017 3:45:16 GMT -5 nick64 said:

May 26, 2017 18:52:43 GMT -5 fhas said:

I think positions are very important. It would be unfair if Cheap Thrills received the same points as #1 that Mask Off received few weeks ago as #6 and that's what would happen with real points.

Why would that be unfair? If that's the amount of points a song earned, that's how much it earned. "Cheap Thrills" was lucky enough to have nothing do better than it that week, while "Mask Off" happened to have five songs do better than it. Doesn't change the fact that they both did about the same amount. It would be unfair to "Mask Off" to discredit it by making it worth less just because it ranked at #6 instead of #1.

Because people preferred to buy / listen to 5 songs instead of Mask Off. Cheap Thills was the most popular song in July / August, even with few points, it's something special in my opinion. And I don't think that just by balancing the real points they would be able to balance this kind of anomaly.

[fhas]

May 26, 2017 23:49:23 GMT -5 @alec935 said:

Did mask off actually have as many points as cheap thrills did when it hit number one? I'm pretty sure it was like over 5k (prolly 10k) less but I might be wrong

It was just an example. But if I'm not mistaken Mask Off peaked with ~35,000 points. Cheap Thrills' 4th week at #1 had aproximately the same points.

[inverse]

May 26, 2017 17:53:47 GMT -5 @holden said:

May 26, 2017 15:33:16 GMT -5 inverse said:

Also maybe I'm reading the numbers wrong but it looks like you have See You Again at 150000 in 2016 and Sail at 175000 in 2014. If those numbers are right then that seems like an error. I know that the cutoff point for the year end hot 100 varies from year to year but there's no way that the point total to miss the list in 2014 was a whole 25000, easily equivalent to a week in the top 10, less than the number 99 in 2016. In fact that's about equal to what Watch Me did in 2016

I disagree. 
Look at 2016, the year end #100 was less than 160,000 points
But this year, the cut-off line is projected to be above 185,000 even with a 0.95X multiplier.

Increasing or adjustments of streaming and formulas can impose a big change to the bottom of a year-end list. That's why I weighed them by case. In 2011 the year-end #100 song was above 200,000 points after the first general weighing. But that's why I also fine-tuned them.

So how is it that after the fine tuning and adjustment Sail still ends up ahead by 25000? With no adjustment at all, assuming Daniel and George perfectly nailed the points of each song, Sail had ~176000 and See You Again had ~149000, but adjusting for formula changes shouldn't they have similar amounts?

[Deleted]

May 27, 2017 16:40:45 GMT -5 inverse said:

May 26, 2017 17:53:47 GMT -5 @holden said:

I disagree. 
Look at 2016, the year end #100 was less than 160,000 points
But this year, the cut-off line is projected to be above 185,000 even with a 0.95X multiplier.

Increasing or adjustments of streaming and formulas can impose a big change to the bottom of a year-end list. That's why I weighed them by case. In 2011 the year-end #100 song was above 200,000 points after the first general weighing. But that's why I also fine-tuned them.

So how is it that after the fine tuning and adjustment Sail still ends up ahead by 25000? With no adjustment at all, assuming Daniel and George perfectly nailed the points of each song, Sail had ~176000 and See You Again had ~149000, but adjusting for formula changes shouldn't they have similar amounts?

That's because ONE adjustment is NOT enough! As I typed in my last update, there's a general multiplier, and one fine-tuning multiplier. I already talked about the structural problem of year-end lists.
Like, for example in year A before streaming era
Raw points for year-end list:
#1 700,000 (X1.5 1,050,000)
#2 675,000
#3 650,000
#5 600,000 (X1.5 900,000)
#10 500,000 (X1.5 750,000)
#20 390,000 (X 1.5 585,000)
#50 250,000 (X1.5 375,000)
#100 150,000 (X1.5 225,000)

Year B in streaming era
Raw points for year-end list:
#1 1,200,000
#2 1,180,000
#3 1,050,000
#5 900,000
#10 750,000
#20 580,000
#50 300,000
#100 150,000

Let me ask you, what multiplier do you apply for year A such that when you compile a decade-end chart or whatever, it's gonna look even? I would use something like X1.5 since when you do decade-end, the top-of-the-year songs are likely effected other than the bottom and that's why the general multiplier is in the top songs' favor.

Now as you noticed, after that general multiplier, the songs at the top are generally even, however, year-end no.1 for year A was still at a disadvantage and THAT'S BECAUSE IN COMPARISON TO THE AVERAGE POINTS OF YEAR-END TOP 100, THE #1 WAS MADE HARDER TO STAND OUT BECAUSE OF PRE-STREAMING ERA you know what I mean. And then, the year-end #100 for year A was 225,000 while for year B was 150,000.

Further to proof that one multiplier/ adjustment per year is definitely NOT enough. That's why we fine-tune songs' points by points and all. But now this is the TOP OF THE DECADE list, so I'm not gonna fine-tune a year-end #100 song.

[Deleted]

Speaking of Sail/See You Again situation, I know that See You Again had year end #3 and #99, but when I apply multiplier and fine-tune, I DID THEM ON A SONG'S ENTIRE POINT TOTAL. So that it doesn't matter if it got #99 on second year or #999, I adjusted its huge points total to balance out with other eras and everything's looking good.

Of course, in one era, there's high points and low points (Uptown Funk vs. Love Yourself). That's why in a generally even era per se (2008-2011) I tend to use one standard system of multiplier on everything, but through 2012 to 2017 every single year there's a change of pace or recurrent rule or whatever. You see in 2014 streaming was far less prominent the general multiplier was higher to match other years and that's probably why Sail got like 175,000 points or whatever and didn't get in. And fast forward to 2015/2016 when the streaming boomed and multiplier cooled down the year end #100 was lower because of top songs' domination and multiplier and all. But in 2017 the huge on-demand data just made the lower tier of year-end more robust meaning an average weekly #30, #50, #100 song had considerably more points in relation to the top songs compared to say, 2015. And that's why this year's year-end #100 is curved up again.

Dude, it's more complicated than you thought. That's why it must've driven most people crazy how to compile a fair list comparing songs from top to bottom from era to era.

[inverse]

^ Ok that makes sense. Sorry for the sort of amateur question.

Also, was there a formula change in 2011 halfway through or was the start of the year just way more competitive, because there were a few occurrences on that year end list that didn't make much sense from a purely chart position standpoint.

[Deleted]

May 27, 2017 17:43:43 GMT -5 inverse said:

^ Ok that makes sense. Sorry for the sort of amateur question.

Also, was there a formula change in 2011 halfway through or was the start of the year just way more competitive, because there were a few occurrences on that year end list that didn't make much sense from a purely chart position standpoint.

Yeah... if you're talking about pure chart position then every year pretty much there are moments like Mask Off at #5 receives more points than Cheap Thrills does at #1. But I didn't specifically dive into that case by case since it's hard to even find billboard hot 100 top 10 articles before 2012 for whatever reason.

But speaking of year end lists that really have some unfair weighing or placements then try 2007, 2012 or 2013. Billboard gave too much weight to the songs in the first few months of 2007 and too little weight to the songs at the end of 2007 (meaning, the big digital/radio formula change, duh). And for 2012 songs mainly charted before March were screwed (like, We Found Love, Set Fire to the Tain, etc) since I remember billboard counted the on-demand streaming data of every song's run before March 24th, 2012 a perfect zero pitted against the songs that released later in the year that have a lot of on-demand advantage). I think they pulled a similar thing for 2013 year-end too or else it's hard to explain why Locked Out of Heaven's relegated to #11.

But of course, if you're trying to hold data together for a decade-end it's good to comply with year-end charts since they're such good sources for it.

[inverse]

I think the ultimate display of complete ridiculousness (from the 2010's, not counting 2007) came in 2011 though.

Are You Gonna Kiss Me Or Not: 92-87-80-69-62-51-42-45-53-58-40-36-33-32-35-34-36-43-47-53. Charted week 5 to week 24. Year end position is 96.

You And Tequila: 92-71-57-50-47-42-37-35-33-35-33-33-33-33-35-35-36-39-50-70. Charted week 28 to week 47. Year end position is 98?!?!?!

Also I think even 2016 had a weird formula change mid-year where streaming numbers were decreased in favor of sales which lead to the 28-39 collapse of Watch Me that week.

[Deleted]

May 27, 2017 19:49:46 GMT -5 inverse said:

I think the ultimate display of complete ridiculousness (from the 2010's, not counting 2007) came in 2011 though.

Are You Gonna Kiss Me Or Not: 92-87-80-69-62-51-42-45-53-58-40-36-33-32-35-34-36-43-47-53. Charted week 5 to week 24. Year end position is 96.

You And Tequila: 92-71-57-50-47-42-37-35-33-35-33-33-33-33-35-35-36-39-50-70. Charted week 28 to week 47. Year end position is 98?!?!?!

Also I think even 2016 had a weird formula change mid-year where streaming numbers were decreased in favor of sales which lead to the 28-39 collapse of Watch Me that week.

Thanks for noticing that!!! In 2011 indeed songs early in the year were at an advantage and the last few months of chart year were screwed. No wonder why Someone Like You was only #24 despite its stellar performance in 2011 chart year.
Songs like Someone Like you, Moves Like Jagger, We Found Love or even Party Rock Anthem were so fucking screwed meaning that 2011 year end was not in favor of them charted late in the year and 2012 year end was not in favor either for them charted early in the year. But just to show how strong PRA was even after double screw still #2 of the decade here!

[Baby Yoda Hot100Fan]

May 27, 2017 19:49:46 GMT -5 inverse said:

Also I think even 2016 had a weird formula change mid-year where streaming numbers were decreased in favor of sales which lead to the 28-39 collapse of Watch Me that week.

They actually rebalanced the Hot 100 formula to reduce the effect of streaming that week, which affected the mid-February chart. They did something similar this year. I don't think it was weird, but much needed, given the continued growth of streaming and decrease in sales. We can probably expect a yearly revision of the formula.

[Deleted]

May 28, 2017 21:44:48 GMT -5 Baby Yoda Hot100Fan said:

May 27, 2017 19:49:46 GMT -5 inverse said:

Also I think even 2016 had a weird formula change mid-year where streaming numbers were decreased in favor of sales which lead to the 28-39 collapse of Watch Me that week.

They actually rebalanced the Hot 100 formula to reduce the effect of streaming that week, which affected the mid-February chart. They did something similar this year. I don't think it was weird, but much needed, given the continued growth of streaming and decrease in sales. We can probably expect a yearly revision of the formula.

Did they pull off a huge rebalance of the formula in 2011 too? 

[Baby Yoda Hot100Fan]

^I'm not sure. But starting with the March 24, 2012 chart, On-Demand Streaming chart was introduced to the Hot 100 formula. Before that, it was radio, digital track sales and some minor streaming counted towards it. That change included Spotify data for the first time. Whether the Hot 100 formula was rebalanced before that I'm not sure.

[Deleted]

been thinking about the fact that BB uses inverse point instead of this. It makes absolutely no sense. The all time list is all luck. It should be based on real points, like this.

[Deleted]

BIG UPDATE (6/17/2017)

I adjusted the multipliers, points system, points and everything from 2010-2017. The biggest update so far!


LEGEND:
**Song Title** = Currently charting
Red = The major year of the song's success
2009-2017 = WEIGHED real points from each year

Year   Basic Multipliers   Approximate Hot 100 Formula                            Approximate sales/airplay/streaming ratio
2010 *1.55                 sales/1, airplay/1
2011 *1.55                 sales/1, airplay/1
2012 *1.25                 sales/1.2, airplay/0.9, on-demand/0.01-0.013              45%, 30%, 25%
2013 *1.15                 sales/1.3, airplay/0.95, streaming/0.05                        34%, 34%, 32%
2014 *1                     sales/1.2, airplay/0.9, streaming/0.06                          32%, 31%, 37%
2015 *1                     sales/1.1, airplay/0.9, streaming/0.085                        30%, 29%, 41%
2016 *1                     sales/0.9, airplay/0.9, streaming/0.12                          29%, 30%, 41%
2017 *0.9                  sales/0.75, airplay/0.9, streaming/0.16                         20%, 25%, 55%

For songs that charted 2008-2011,
Points                 Fine-tuning Multiplier =
1,500,000+                            1.05
1,400,000+                            1.04
1,300,000+                            1.03
1,250,000+                            1.02
1,200,000+                            1.01

(00S TOTAL + 10S TOTAL does NOT necessarily add up to TOTAL, since everything's based off of song's performance in a certain time period earning different basic and fine-tuning multipliers)

[Deleted]

2010S DECADE-END HOT 100 SONGS (REAL POINTS VERSION) (AS OF 6/17/17)

# Title Artist
1 UPTOWN FUNK!, Mark Ronson Featuring Bruno Mars
2 PARTY ROCK ANTHEM, LMFAO Featuring Lauren Bennett & GoonRock
3 ROLLING IN THE DEEP, Adele
4 SOMEBODY THAT I USED TO KNOW, Gotye Featuring Kimbra
5 BLURRED LINES, Robin Thicke Featuring T.I. + Pharrell
6 CLOSER, The Chainsmokers Featuring Halsey
7 RADIOACTIVE, Imagine Dragons
8 DARK HORSE, Katy Perry Featuring Juicy J
9 THRIFT SHOP, Macklemore & Ryan Lewis Featuring Wanz
10 HAPPY, Pharrell Williams
11 SEE YOU AGAIN, Wiz Khalifa Featuring Charlie Puth
12 CALL ME MAYBE, Carly Rae Jepsen
13 ALL OF ME, John Legend
14 ALL ABOUT THAT BASS, Meghan Trainor
15 SHAKE IT OFF, Taylor Swift
16 JUST THE WAY YOU ARE, Bruno Mars
17 HELLO, Adele
18 MOVES LIKE JAGGER, Maroon 5 Featuring Christina Aguilera
19 COUNTING STARS, OneRepublic
20 ROYALS, Lorde
21 THINKING OUT LOUD, Ed Sheeran
22 WE ARE YOUNG, fun. Featuring Janelle Monae
23 DYNAMITE, Taio Cruz
24 THE HILLS, The Weeknd
25 SORRY, Justin Bieber
26 TRAP QUEEN, Fetty Wap
27 GIVE ME EVERYTHING, Pitbull Featuring Ne-Yo, Afrojack & Nayer
28 FIREWORK, Katy Perry
29 ROAR, Katy Perry
30 WE FOUND LOVE, Rihanna Featuring Calvin Harris
31 TIK TOK, Ke$ha
32 WAKE ME UP!, Avicii
33 CAN'T STOP THE FEELING!, Justin Timberlake
34 F**K YOU (FORGET YOU), Cee Lo Green
35 HARLEM SHAKE, Baauer
36 STAY WITH ME, Sam Smith
37 E.T., Katy Perry Featuring Kanye West
38 LOVE YOURSELF, Justin Bieber
39 ONE DANCE, Drake Featuring WizKid & Kyla
40 NEED YOU NOW, Lady Antebellum
41 GRENADE, Bruno Mars
42 PAYPHONE, Maroon 5 Featuring Wiz Khalifa
43 SEXY AND I KNOW IT, LMFAO
44 WATCH ME, Silento
45 HEY, SOUL SISTER, Train
46 SOME NIGHTS, fun.
47 FANCY, Iggy Azalea Featuring Charli XCX
48 HO HEY, The Lumineers
49 DON'T LET ME DOWN, The Chainsmokers Featuring Daya
50 LOVE THE WAY YOU LIE, Eminem
51 CAN'T FEEL MY FACE, The Weeknd
52 ONE MORE NIGHT, Maroon 5
53 WRECKING BALL, Miley Cyrus
54 LIGHTS, Ellie Goulding
55 CALIFORNIA GURLS, Katy Perry
56 BLANK SPACE, Taylor Swift
57 STRESSED OUT, twenty one pilots
58 HOTLINE BLING, Drake
59 SHAPE OF YOU, Ed Sheeran
60 SOMEONE LIKE YOU, Adele
61 WORK, Rihanna Featuring Drake
62 CHEAP THRILLS, Sia Featuring Sean Paul
63 SUGAR, Maroon 5
64 PANDA, Desiigne
65 RUDE, MAGIC!
66 STITCHES, Shawn Mendes
67 CRUISE, Florida Georgia Line Featuring Nelly
68 OMG, Usher Featuring will.i.am
69 LOCKED OUT OF HEAVEN, Bruno Mars
70 CAN'T HOLD US, Macklemore & Ryan Lewis Featuring Ray Dalton
71 SHUT UP AND DANCE, WALK THE MOON
72 BAD ROMANCE, Lady Gaga
73 AIRPLANES, B.o.B Featuring Hayley Williams
74 STRONGER (WHAT DOESN'T KILL YOU), Kelly Clarkson
75 TEENAGE DREAM, Katy Perry
76 SUPER BASS, Nicki Minaj
77 GLAD YOU CAME, The Wanted
78 TIMBER, Pitbull Featuring Ke$ha
79 HEATHENS, twenty one pilots
80 MIRRORS, Justin Timberlake
81 PUMPED UP KICKS, Foster The People
82 DJ GOT US FALLIN' IN LOVE, Usher Featuring Pitbull
83 TAKE ME TO CHURCH, Hozier
84 JUST GIVE ME A REASON, P!nk Featuring Nate Ruess
85 TALK DIRTY, Jason Derulo Featuring 2 Chainz
86 CHEERLEADER, OMI
87 EARNED IT (FIFTY SHADES OF GREY), The Weeknd
88 679, Fetty Wap Featuring Remy Boyz
89 I LIKE IT, Enrique Iglesias Featuring Pitbull
90 WHAT DO YOU MEAN?, Justin Bieber
91 DEMONS, Imagine Dragons
92 LEAN ON, Major Lazer & DJ Snake Featuring M0
93 POMPEII, Bastille
94 STARBOY, The Weeknd Featuring Daft Punk
95 WHEN I WAS YOUR MAN, Bruno Mars
96 PROBLEM, Ariana Grande Featuring Iggy Azalea
97 SET FIRE TO THE RAIN, Adele
98 THE MONSTER, Eminem Featuring Rihanna
99 BREAK YOUR HEART, Taio Cruz Featuring Ludacris
100 STARSHIPS, Nicki Minaj

[Deleted]

May 28, 2017 17:35:26 GMT -5 @holden said:

May 27, 2017 19:49:46 GMT -5 inverse said:

I think the ultimate display of complete ridiculousness (from the 2010's, not counting 2007) came in 2011 though.

Are You Gonna Kiss Me Or Not: 92-87-80-69-62-51-42-45-53-58-40-36-33-32-35-34-36-43-47-53. Charted week 5 to week 24. Year end position is 96.

You And Tequila: 92-71-57-50-47-42-37-35-33-35-33-33-33-33-35-35-36-39-50-70. Charted week 28 to week 47. Year end position is 98?!?!?!

Also I think even 2016 had a weird formula change mid-year where streaming numbers were decreased in favor of sales which lead to the 28-39 collapse of Watch Me that week.

Thanks for noticing that!!! In 2011 indeed songs early in the year were at an advantage and the last few months of chart year were screwed. No wonder why Someone Like You was only #24 despite its stellar performance in 2011 chart year.
Songs like Someone Like you, Moves Like Jagger, We Found Love or even Party Rock Anthem were so f**king screwed meaning that 2011 year end was not in favor of them charted late in the year and 2012 year end was not in favor either for them charted early in the year. But just to show how strong PRA was even after double screw still #2 of the decade here!

How was PRA screwed? It ranks higher in BBs list than yours. Points don't matter to them, and the all-time charts are what is gonna go down in history.

I think the song really screwed is Somebody That I Used To Know. Peaked ridiculously high in he beginning, and was an extremely massive song, but didn't even hit double digit weeks. One dance is a much bigger song according to BB (I think.)