Rank and sales data could help figure out App Store popularity algorithm
Joel Comm, publisher of the highly-useful app iFart Mobile, has published some data about how sales related to popularity in the App Store.
While I have only a high-school understanding of statistics, it would appear from Comm's data that as sales rise in a linear fashion, App Store ranking increases (or I should say decreases, because lower numbers are better) logarithmically.
Comm admits that his data isn't carefully tabulated: he hand-counted the sales, as they're broken down by country, and says he didn't include countries "where one or two units were sold." Comm's data was collected between December 12th and 17th.
On Sunday's talkcast, the Iconfactory's Craig Hockenberry and Gedeon Maheux lamented the lack of analysis tools for sales, and said they'd appreciate more transparency about how rankings are mathematically determined.
Comm anticipates more data will be forthcoming from other developers. With more data, a more detailed statistical analysis can be performed, and preferably from someone other than a TUAW blogger who majored in art.
