EvanJ
6x Platinum Member
Joined: September 2003
Posts: 6,371
|
Post by EvanJ on Jan 30, 2014 18:31:20 GMT -5
I looked at 640 charts from the first available Hits Daily Double chart on April 15, 2000 through January 17, 2014. I wanted to measure change over time and I decided to exclude all charts in December when albums sell the most along with weeks when Hits Daily Double didn't have a chart available. The Top 10 combined ranged from 307,614 to 3,638,908 with a mean of 1,060,410. The Number 1s ranged from 41,057 to 1,977,786 with a mean of 291,411, and almost two-thirds of the Number 1s sold below the mean. The top week of any album (1,977,786 for "Celebrity" by *NSYNC) sold more than the Top 10 combined in exactly 95 percent of the weeks I looked at. The correlation between week number (from 1 to 640) and Top 10 sales was -0.5991. The correlation between week number and Number 1 sales was -0.2763. Both of these show that sales have decreased over time. I made a big spreadsheet with four tabs and I will e-mail it to anybody who is interested.
|
|
|
Post by ListenToItTwice on Jan 30, 2014 23:50:38 GMT -5
Did you do any comparison of Hits and Soundscan Top 10s?
|
|
EvanJ
6x Platinum Member
Joined: September 2003
Posts: 6,371
|
Post by EvanJ on Jan 31, 2014 19:11:47 GMT -5
Did you do any comparison of Hits and Soundscan Top 10s? No, because Hits Daily Double is the only source I know of with exact numbers. I didn't want numbers rounded to the nearest thousand.
|
|