Chris McEvoy, the man behind the Userati, has created an interesting new mashup called Book Rank. Here's how it works:
- Take a single book from Amazon as a root.
- Take the recommended books for that book.
- Repeat until you have reached a distance of 20 from the original book.
- Add up the number of recommendations for each book.
- Calculate BookRank from the Distance, Recommendations, SalesRank.
- Display the 250 books with the highest BookRank.
Seems like a useful, novel way to find books.



Something I've been wondering about these recommendations: is there a way to know whether they're "up" or "down"? I.e., is Amazon recommending something obscure based on a popular book, or something popular based on an obscure one?