BILLBOARD 2010s ARTIST OF THE DECADE
READING TIME: 15 MINUTES
INTRODUCTION AND BACKGROUND
If one has to think of the music titans who ruled the 2010s one might consider who dominated on radio, or who had stellar sales or even a huge streaming presence; they might even consider other factors such as touring and a big social media presence. Names of those who stomped on these multiple formats might include Rihanna, Drake, Beyonce, Nicki Minaj, Justin Bieber - just to name a few. In the United States, week by week, Billboard Magazine publishes the most popular songs on the Hot 100 for singles and the Billboard 200 for albums. They also keep a tally of how each release performs and by the end of the decade Billboard releases a list of who are the most popular recording artists of the decade. This is done by aggregating the artist’s total points. For the 2010s the chart points were for the period between December 2009 and November 2019. The previous winners and runners up were Eminem and Usher for the 2000s and Mariah Carey and Janet Jackson for the 1990s. These artists had consistent commercial success throughout their relevant decades.
On October 14th 2020, the 2010s Billboard Artist of the Decade winner will be revealed at the 2020 Billboard Music Awards. The positions for the Artists of the Decade from position #100 down to #6 have already been revealed. However, the top 5 have not. It is speculated that they are Bruno Mars, Taylor Swift, Drake, Adele and Rihanna, but not in that order. It is possible to predict who the winner is by using Billboard’s previous methods to determine the artist of the decade. It is speculated that in the 1990s the list was determined using the total weekly charting points. The weekly points were a raw number of points of how much a recording was consumed both physically and on the radio. On the other hand, for the 2000s Billboard determined their list by using an inverse point system. This is where the releases’ weekly placement was tallied and used to calculate their decade-end performance instead. So a #1 album or single on a given week would have 100 points, a #2 release has 99 points, and so on. For the 2010s, publications believe that Billboard will follow a similar format as the 2000s. This is because it is incredibly difficult to keep track of points with multiple types of streaming and radio formats. The ease of streaming, in comparison to sales, has lead to longer chart runs and also makes it unfair for songs that debuted earlier without streaming. Therefore it is also believed that songs before 2015 (at the point when streaming was showing its true power) will have a higher weighting compared to those released in the latter half of the decade.
Sometimes it is possible to predict who the top artists are by their looking their positions on the already released lists for the Songs of the Decade and Albums of the Decade. Therefore this investigation will try to determine who the Artist of the Decade is using these two lists.
*I WOULD LIKE TO GIVE THANKS TO SARAH ZENG (https://www.linkedin.com/in/sarahzeng/) FOR HER ADVICE IN THIS PREDICTIVE RANK INVESTIGATION*
METHDOLOGY
Singles and Albums Performance: I went through the Songs of the Decade and Albums of the Decade lists and made my own list of artists who made multiple appearances on each list. Then I wrote down the positions the artist’s releases were placed. Then, to quantify the performance of each release that charted I used the inverse point system. This is because there are too many factors to take into account for consuming music in the 2010s, and the inverse point system is the simplest method. Unfortunately, it means that an album such as ‘25’ which debuted with 3,380,000 copies, had the same weekly chart performance figure as ‘Anti’ which debuted with 166,000 copies. Using this methodology as an example, the #1 Song of the Decade was ‘Uptown Funk’ by Mark Ronson and Bruno Mars. Therefore I awarded it 100 points. I split the points 70% to Mark Ronson and 30% to Bruno as this is the general rule Billboard follows when splitting a main artist’s versus a featured artist’s points. As there are 200 placements on the Billboard 200 compared to 100 placements on the Billboard Hot 100 lists, I halved the points for each BB200 album placement. This is because I gave the Hot 100 and the BB200 equal weighting, as successful artists have always had both strong singles and albums success. For example, the #33 Album of the Decade was the ‘A Star is Born (Soundtrack)’ by Lady Gaga and Bradley Cooper. Therefore I originally awarded it 78 points and then I halved it to give it 39 points. Thereafter I halved the amount of points once more as Lady Gaga and Bradley Cooper were both named lead artists so they both achieved 19.5 points for this project. I carried out this method for every release that charted. Then I ordered the lists to see who had the best performing Hot 100 and BB200 success on the chart. I also added together the artist’s Hot 100 and BB200 points to get to an overall Artist of the Decade rank. I also created graphs to show how the releases of the predicted Top 10 Artists did annually and cumulatively over the decade. I then went back and added an SPS (streaming plus sales) weighting factor. Here I added a multiplier of 0.5x to songs released from 2015 onwards as releases began to benefit from streaming. Streaming has been shown to elongate chart runs compared to songs released earlier in the decade, which would give these songs an unfair advantage. This is because streaming allows a release to be continuously consumed, more than it would have done just from single purchases on sales platforms or the radio. This multiplier was the fairest way to take streaming into account when creating these lists. Therefore the ratio/strength of pre-2015 and post-2014 songs is 1.0 : 1.5.
Label Performance: I went through my predicted Top 25 Artists of the Decade and I wrote down the label which the artist’s projects were signed to. Some artists such as Ed Sheeran released projects under two labels (Atlantic and Asylum) throughout the decade so this was also taken into consideration. Then I worked out the proportion of artists who were signed under the major labels Universal Music Group (UMG), SONY Music Entertainment, Warner Bros Music and BMG. I then worked out the proportion of artists who were assigned to their specific label, under each parent group. From this I created a visual to show the presence of the most popular and least popular labels, based on the Top 25 Artists.
Male vs Female Performance: Using the predicted Top 10 Artists of the Decades, I used the top male and top female artist’s performance to work out if the males or the females performed better every year during the 2010s.
Singles vs. Albums Artist: I wanted to see, based on the collated data, whether the predicted Top 10 Artists of the Decade garnered their chart success through their singles or their albums. Therefore I worked out what proportion of their performance was built up by singles success and what was built up by their albums success. It can help build up a picture to see if an artist is a ‘singles artist’, an ‘albums artist’ or both (which is usually the ingredient for the very top artists of the decade). The limitation of this investigation is that this does not fully determine their status as an albums/singles artist as it only takes into consideration the Top 100/200 lists. Billboard will have songs below the Top 100/200 which might changes the artist’s proportions. However, the Top 100/200 will carry more weight because they are bigger releases and should be a good base determinant.
Accuracy of Ranking Models: I was interested to see how accurate my predicted Artist of the Decade rank was compared to Billboard’s rank (using the unofficial leaked list). Therefore I used a Kendall Tau software (Wessa, (2017), Kendall tau Rank Correlation (v1.0.13) in Free Statistics Software (v1.2.1), Office for Research Development and Education, URL https://www.wessa.net/rwasp_kendall.wasp/) which would help identify the statistical association between my ranking and Billboard’s ranking. As I was creating a bivariate correlation Kendall Tau could quantify the strength of the relationship. I replicated the same entire above methodology for the 90s and 00s Top 10 Artists of the Decade. It would help show how accurate my methodology is compared to the actual list. I chose the Kendall Tau ranking method because it is a non-parametric test meaning that does not depend on assumptions of various underlying distributions. The method is also beneficial when handling small figures and smaller sample sizes (n=10) compared to other ranking systems such as Spearman’s rho. The interpretation of Kendall Tau in terms of the probabilities of observing concordant (agreeable) and discordant (non-agreeable) pairs is direct. The null hypothesis was that there is an association between my predictive ranking and the actual ranking. It should be noted that the main limitation of my ranking is that Billboard’s list also takes into account the entire chart performance of the artists, not just the Top 100/200 lists used for this investigation.
RESULTS & DISCUSSION
Singles Performance (of songs placed in the Top 100):
Figures 1 (L) and 2 (R): Line charts that show the cumulative performance of the predicted Top 10 Decade End Artists’ songs which reached the Top 100 Decade End. The left graph shows the performance using raw points whereas the right graph shows the points with SPS-weighting, which is why the y-axis decreases by 50 points. The dotted line splits the pre-streaming and post-streaming eras.
Figure 1 shows that there was a competitive singles market in 2010/11 with Adele, Katy Perry and Bruno Mars releasing commercially successful singles such as ‘Rolling in the Deep’ (#10), ‘ET’ (#40) and ‘Just the Way You Are’ (#22). Maroon 5 slowly caught up with nearly 150 points in 2012 due to ‘Moves Like Jagger’ (#20) and ‘One More Night’ (#28). Despite Rihanna achieving 7 #1 songs between 2010 and 2012, only ‘We Found Love’ (#6) garnered enough points to compete with her peers, and so she fared much farther behind. Katy Perry and Bruno Mars continued their singles dominance with their upcoming projects from ‘Unorthodox Jukebox’ and ‘Prism’ which contained the singles ‘Locked Out of Heaven’ (#45) and ‘Dark Horse’ (#26). By the middle of the decade Katy Perry dominated the singles market with 5 entries, followed by Bruno Mars, Adele and Maroon 5. During the second half of the decade Bruno Mars superseded Katy Perry with Song of the Decade ‘Uptown Funk’ and the singles from ‘24K Magic’ leaving him as this investigation’s Hot 100 Artist. Also more male artists began seeing success in the singles market as Drake, Ed Sheeran, Justin Bieber and Post Malone who began to have multiple successful singles such as ‘God’s Plan’ (#19) , ‘Shape of You’ (#3), ‘Love Yourself’ (#42) and ‘Sunflower’ (#11). This helped push them above Rihanna and Taylor Swift by the end of the decade. It is interesting to note that Post Malone, who on this graph debuted last in 2017, has the greatest Hot 100 rise out of all his peers achieving approximately 280 chart points by the end of the decade. Using Figure 2, once you take SPS-weighting into account these male artists who dominated in the second half of the decade had suppressed chart points and it becomes clear that Ed Sheeran had the strongest singles longevity, followed by Post Malone. Bruno Mars, Katy Perry and Adele are fortunate that their placings are largely unaffected as they achieved most of their Hot 100 smash single success before the streaming era.
Albums Performance (of albums placed in the Top 200):
Figures 3 (L) and 4 (R): Line charts that show the cumulative performance of the Top 10 Decade End Artists’ albums which reached the Top 200 Decade End. The left graph shows the performance using raw points whereas the right graph shows the points with SPS-weighting, so the y-axis decreases by 100 points. The dotted line splits the pre-streaming and post-streaming eras. The x-axis should follow FIgures 1 and 2 and contain the years 2010 to 2019.
Figure 3 shows that the albums market was oversaturated from 2010 to 2013. Drake, Taylor Swift, Adele and Justin Bieber released successful albums such as ‘Take Care’ (#45), ‘Red’ (#32), ‘21’ (#1) and ‘My World 2.0’ (#22) in the digital age. Bruno Mars, Rihanna, Katy Perry and Maroon 5 still had albums success but had a more gradual rise in comparison to their aforementioned peers. By the middle of the decade Drake and Taylor Swift were almost neck-and-neck for being the most consumed albums artist with almost 3 projects each to their name. These two artists had also garnered nearly as many points as their singles performance across the entire decade showing that they received most of their charting points from the BB200 success. As we move into the second half of the decade, Taylor Swift benefitted from 2 more highly successful albums including 1989 (#2). However, Drake’s album success was unparalleled and his two projects Views (#9) and Scorpion (#19) pushed him to earning more than double the points compared to his peers. It is also important to note that his latter albums benefitted from being released in the streaming era which allowed him to have a longer chart run. Drake’s performance showcases a continuous climb throughout the decade with so many releases under his belt compared to his peers who released less frequently. Similar to Figure 1 and 2, Ed Sheeran and Post Malone also had immense albums success with ‘Divide’ (#3) and ‘Beerbongs & Bentleys’ (#5). Using Figure 4, the SPS-weighting suppresses everybody’s chart performance. Drake falls by almost 12% compared to Taylor Swift who only suffered a 9% decrease in chart performance, yet it is clearly defined that Drake and Taylor are the albums artists of the decade.
Aggregated Cumulative Performance:
Figures 5 (L) and 6 (R): Line charts that show the cumulative performance of the predicted Top 10 Decade End Artists’ singles and albums together. The left graph shows the performance using raw points whereas the right graph shows the points with an SPS-weighting.
Similar to Figure 3, the 2010-2013 music market was extremely competitive with Adele, Bruno Mars and Katy Perry dominating the Billboard charts with ‘21’ (#1), ‘Unorthodox Jukebox’ (#15) and “Teenage Dream’ (#44). Maroon 5, Rihanna, Drake, and Taylor Swift were establishing themselves but at a much slower rate during the first half of the decade. They were most likely experiencing success at this time in one platform as opposed to both. By the end of the first half of the decade it appears Katy Perry was on good track to be Artist of the Decade if she was to keep up the pace of her chart success. Unfortunately, she remained stagnant with no major smash singles or albums to help contribute to growing her chart success. Drake, Taylor Swift and Adele released further projects, such as ‘25’ (#19), ‘Blank Space’ (#65) and ‘One Dance’ (#35), to push themselves above their peers and into the top echelons.. Bruno Mars’ 24K Magic era put him on par with Drake by 2017. However, Scorpion and its singles gave Drake that final push to end up with the Artist of the Decade title. Rihanna and Justin Bieber experienced slow, consistent success with their projects ‘Anti’ (#29) and ‘Despacito (Remix)’ (#9) which helped give them mid-tier placements. Ed Sheeran and Post Malone then crept in with enormous success in the second half of the decade and were able to squeeze themselves into comfortable mid-to-lower tier positions of the Top 10. Using SPS-weighting in Figure 6, Drake’s points are suppressed by almost 25% showing how much he benefited from the streaming era. Meanwhile, the first half of the decade appears less deflated and it becomes obvious that Adele, Bruno Mars and Taylor Swift were experiencing a continuous upward climb establishing themselves as strong candidates for being in the upper realm for Artist of the Decade. The Drake domination in Figure 5 becomes less impressive in Figure 6 once you see that there is a more equal spreading of the artists’ cumulative performance as opposed to Drake demolishing the competition. From Figure 6 you can infer that the Artists of the Decade rank is Drake (#1) followed by Bruno Mars, Taylor Swift, Adele, Katy Perry, Maroon 5, Justin Bieber, Ed Sheeran, Rihanna and Post Malone.
Aggregated Annual Performance:
Figures 7 (upper) and 8 (lower): Line charts that show the annual performance of the Top 10 Decade End Artists’ singles and albums together. The left graph shows the performance using raw points whereas the right graph shows the performance with a 1.0 : 1.5 weighting non-SPS: SPS. The yellow arrows indicate deflated chart points from an average of 200 chart points in Figure 7 to 140 chart points in Figure 8 (30% drop).
Figure 7 shows that Adele, Bruno Mars and Katy Perry gained the highest charting points in 2011. Adele achieved the highest peak of the decade with her project ‘21’ with ‘Rolling in the Deep’ carrying most of her points. Both of these releases gave her Top 10 End of Decade Placements in the Hot 100 and BB200 charts. Taylor Swift, Drake and Rihanna showed smaller grains of success but they were more consistent with their releases throughout the decade. The music market appears weakest during the middle of the decade with only 2 artists displaying major success. One possible cause of this was most the artists in the first half of the decade were either oversatured or exhausted with their releases, which left the middle of the decade open to new up-and-coming artists instead - this is where Ariana Grande and The Weeknd began establishing themselves.. Again Figure 7 seems that the second half of the decade appears more competitive than the first. Drake, Justin Bieber, Ed Sheeran and Post Malone experienced the peaks of their careers within years of one another. Drake is fortunate to have experienced two peaks in 2016 and again in 2018. However, it was more albums driven in 2016 and more singles driven in 2018. Without taking SPS into account (ie. Figure 7) one could interpret that their eras were experiencing almost 80% of the success Adele achieved during 2011. However, when looking at Figure 8 and taking SPS-weighting into account, the second half of the decade becomes more depressed and the peaks are less steep. They are more on par with ‘Unorthodox Jukebox’ and ‘Overexposed’ which is more realistic as ‘21’ was a rare experience of epic proportions at the time in the industry. Artists highly successful in the first half of the decade also made strong appearances in the second half of the decade, such as Rihanna, Adele, Bruno and Taylor - all of whom were establishing consistent commercial success through the decade.
Label Performance
Figure 9: Stacked bar charts which show the breakdown for how much the Big 4 Parent Music Labels dominated the decade. The %s show how many artists/projects were released under each label during the 2010s.
Figure 9 shows that UMG signed the most successful artists to its parent label because 68% of the Top 25 Artist releases came from UMG. This probability matched the predicted Top 10 Artists of the Decade as 70% of the artists had also been signed to UMG during the decade. Republic and Interscope held 40% of the top UMG artists in its roster with names such as Drake, Post Malone, Maroon 5 and Lady Gaga. Sony experienced 70% less artist presence in comparison to UMG. Their most successful flagship labels were Columbia who had Beyonce and Adele signed to them and RCA who had P!nk and Justin Timberlake. Warner Music shared 6% of the Top 25 artists largely benefiting from having Ed Sheeran and Bruno Mars. Their presence is 60% less than Sony and 88% less than UMG, again demonstrating that UMG are the biggest and strongest label in the past decade. Finally BMG held 4% of artists releases during the decade. Taking all of this into account, it can be inferred that if one is hoping for musical commercial success your best bet is to sign to UMG as they have had the greatest presence amongst the top recording artists of the decade.
Male vs Female Performance (only taking into account the Predicted Top 10):
Figures 10 (L) and 11 (R): Line charts that show the male vs female performance of the Top 10 Decade End Artists. The left graph shows the performance using raw points whereas the right graph shows the raw inverse points performance but with a 0.5x multiplier in the second half of the decade to take into account the benefit of streaming. The black dotted line shows the total accumulated points, which is the trendline for the overall performance across the decade.
Figure 10 shows that at the beginning of the decade the biggest chart performing artists on Billboard were female such as Adele, Taylor Swift and Katy Perry. The market was viciously competitive and this is shown by the peak of over 800 points witnessed in 2011. It is likely the market benefitted from Adele’s ‘21’ project in order to reach this peak. Then in 2012 and 2013 the market appears weakened and male artists began to surpass female artists, with the biggest contributions coming from Drake, Bruno Mars. and Maroon 5 By 2014 no popular male artist had released a commercially successful project in order to continue their dominance. As a result the top performing female artists, Katy Perry and Taylor Swift, continued female commercial success across 2013 to 2014. By 2014 it appears the music market was at its weakest showcasing a 75% drop compared to 2011. In 2015, Taylor Swift was the only major female artist who was able to sustain a female presence. From this point male artists dominated the market and the black dotted line follows the males trend line more closely than the females, to also establish their dominance. For example, in 2018 the ratio of M:F commercial performance is 5.5 : 1.0. They were probably aided by streaming and the rise of hip-hop and tropical-house which are male dominated genres. Artists also realised that you have to release more projects to achieve commercial success in the streaming era which gave the males an advantage to garnering more chart points. Figure 11 shows a depressed version of the second half of the decade taking streaming into account. The female trend line also becomes a further exaggerated decline showing the lack of female presence felt in the industry during those years only kept afloat by Taylor Swift, Adele and Rihanna.. The depression also helps show that the males’ success is not as high as what the females experienced in 2011 which Figure 7 would lead the reader to believe. The limitation of this graph is that the predicted Top 10 Artists of the Decade has only 4 female artists and 6 male artists (including groups) which can skew black line in favour of the males and not giving an equal reflection of M:F performance.
Singles vs. Albums Artist
Figures 12 (upper) and 13 (lower): A series of pie charts which shows the breakdown of how much an artist’s total points is made up using their singles performance and of their albums performance. The top 10 charts are without the SPS adjustment whereas the charts below are inclusive of the SPS adjustment. From left to right: Drake, Taylor Swift, Bruno Mars, Rihanna, Adele, Ed Sheeran, Justin Bieber, Katy Perry, Maroon 5 and Post Malone.
Figure 11 shows that Drake, Taylor Swift, Rihanna and Justin Bieber achieved the majority of their top Billboard points from their album performance compared to their singles performance. This is similar to what we saw in Figures 3 and 4 where Drake and Taylor’s cumulative performance is more dominant towards their albums. They both achieved more than 75% of their charting points from their album’s performance. Drake did this with 8 BB200 chart placements and 5 Hot 100 placements and Taylor did the same with 5 and 2 projects, respectively. Rihanna, once defined as a ‘singles artist’, surprisingly achieved 65% of her success from albums over singles. This may be because her credited single’s success was diluted and she had to split the singles’ performance with featured artists such as Eminem, Drake and Calvin Harris. Meanwhile her albums performance is 100% credited just to herself.
On the other hand, Katy Perry and Maroon 5 experienced most of their success with approximately 60% weighting from the success of their singles compared to their albums. Katy Perry had 5 Hot 100 placements compared to only 2 BB200 placements which would heavily weigh her singles performance greater than her albums performance. Similarly Maroon 5 had 4 singles on the Hot 100 list compared to only 3 on the BB200.
The remaining artists Bruno Mars, Adele, Ed Sheeran and Post Malone appear to have equal success in both their albums and their singles. Adele and Post Malone’s majority falls on their BB200 success whereas Ed Sheeran and Bruno Mars’ success is slightly skewed to their singles success.
Figure 12 shows the same artist’s composition but takes the SPS-weighting into account. Drake, Taylor Swift and Justin Bieber’s album composition increased by a few % due to the increased weight from their successful pre-2015 album releases. The remaining artists had slight percentage changes or did not change at all. Katy Perry and Post Malone would not experience any changes as their biggest chart performances were only experienced in one half of the decade.
Accuracy of Ranking Models
Figure 14: Scatterplot graphs using the Kendall Tau correlation coefficient for the 1990s (top-left), the 2000s (top-right), the 2010s without SPS-weighting (bottom-left) and the 2010s with SPS-weighting (bottom-right). The x-axis represent where I rank the artist’s placements and the y-axis represents the actual rank, where Billboard ranks the artist’s placements. Data points that lie above the line indicate that these artist’s ranking are underrepresented using my rank model and data points that low below the line indicate that these artist’s ranking are overrepresented using my rank model. The dotted line in the bottom-left graph indicates the biggest difference in rank placements out of all the artists on all four plots.
Figure 14 shows that the Kendall Tau correlation coefficient, using my predictive rank model, creates a weak positive correlation in the 1990s (+0.56) and a strong positive correlation for the 2000s (+0.82). Then in the 2010s, without taking SPS in account, my predictive rank model again creates a weak positive correlation (+0.56) and a very strong positive correlation once taking SPS into account (+0.87).
The weaker rank models have 1 or 2 artists who are accurately ranked compared to the stronger rank models who have 4 to 6 artists who were accurately ranked. The weaker rank models also have 5 Top 10 artists whose ranking is underrepresented compared to their actual rank and have 3 to 4 Top 10 artists whose ranking is overrepresented compared to their actual rank. Of particular note is Rihanna’s rank in the ‘2010s - SPS’ plot (see the bottom-left plot). My rank predicted her to be in 10th place, which is 6 places away from the predicted actual rank which is the greatest difference of any artist’s rank.
It can be inferred that the inverse point system, which I emulated from the 2000s decade-end list and applying the SPS-weighting, is what helped improved the correlation co-efficient and is the ranking model Billboard actually used to compile their list. As the correlation co-efficient is weaker in the 1990s the method used to create the rank perhaps used more factors than what is used in recent times. Alternatively they used the raw total chart point system as opposed to the inverse point system, as mentioned in the methodology.
CONCLUSION
Figure 15: A table showing the “unofficially released” Billboard Top 10 list and my list both with -SPS and +SPS. The green cells show the artist was accurately placed. The orange shows the artist was placed one rank position away from their actual position. The accuracy rate increases from 60% to 80% once taking SPS into account speculating that the 1.0 : 1.5 was a similar ratio they gave to post-2014 releases.
To round up this investigation I would to note that the music landscape is constantly evolving both in terms of genre and how music is consumed. Therefore different statistical methods and models will have to be taken into consideration in order to quantify the most popular releases in a given week or decade. This is why Billboard, and myself, have used multipliers to try and give a fair balance to different musical eras. Many eras have a different singles and albums turnover rate which can distort the rankings and give unfair disadvantages to songs released in a climate with faster turnover rates.
From my analysis, I predict that Drake will be crowned the Artist of the Decade. This will be the second consecutive decade a rapper has been awarded the title. This may come to no surprise seeing that he consistently released commercially successful projects nearly every single year in the decade. Taylor Swift should become the runner-up and be listed the female artist of the decade, following the footsteps of Beyonce and Madonna in their respective decades. Surprisingly, Bruno Mars, Adele and Ed Sheeran only released 2 or 3 projects throughout the entire decade which showcases their high commercial value by the general public. Nonetheless all the top 10 artists exhbited strong commercial releases in both albums and singles in order to attain their Top 10 positions. The most surprising outcome from this investigation is seeing Post Malone’s Billboard domination after only three years of success in the latter end of the decade - this was enough to land him a Top 10 position! Also Rihanna’s #4 position was probably bolstered by her bubbling-under Top 100 Hot 100 success as neither of my models predicted her to rank so high.
If I continued this investigation further I would use the 1990s model and look at using numerous components to compile this list. I could compare if the inverse point system model or the raw chart points system is more closely related to the leaked rank (as seen in Figure 15). I would also be interested in changing the SPS-weighting throughout the 2010s decade, as opposed to just using a pre-2015 and post-2014 multiplier. It may help me see if this would create a more accurate rank which Billboard lists because streaming, sales and radio underwent major weightings changes to come up with the Hot 100 and BB200 formula. For all we know Billboard may incorporate an unknown ranking system we are not aware of, or they may even take touring and social media interaction data into account.
