Some more info about their readability formula: it uses "the number of words per sentence, the number of characters per word, and the average grade level of the

Some more info about their readability formula: it uses "the number of words per sentence, the number of characters per word, and the average grade level of the words in the book". http://doc.renlearn.com/KMNet/R003520002GE7114.pdf http://doc.renlearn.com/KMNet/R003520002GE7114.pdf Personally I don't think this is a very good marketing strategy, and I'm not surprised it's hidden in a 10MB PDF document four levels deep on their website. Either they should: 1. Go deeper and do things like comparing complexity of parse trees of sentences. This might bring ratings more into line with NYT columnists' expectations, but more importantly, it sends a signal to buyers that "we think about this more deeply than you can understand". (And as an added bonus it would generate jobs for people like me.) 2. Give out points randomly --- unpredictable rewards are better motivation than fixed rewards, and shield you from criticism by people who don't like the fact that their chosen books don't earn enough points, so you can spend more time selling and less time trying to justify your ratings mechanism.