This is a well-defined statistical study of the art historical literature of a particular period. It counts the number of times that ancient artworks are mentioned in Renaissance art literature. By measuring the popularity and co-popularity of artworks it uncovers several interesting facts.
Firstly, canons are identical with the most popular items over a distribution of popularity. Secondly, sub-tails of genres or subjects have broadly the same properties as the main long tail of which they are a sub-tail. And thirdly the co-popularity of otherwise unrelated monuments may be a product of their spatial proximity at the time they were documented in the renaissance.
These facts are interesting in themselves and indicate further possibilities for research. They are also of use to more theoretical or social art historical approaches.
(I originally posted about this here.)