Following up on our earlier stories about the Amazon astroturf effort, and even more particularly on yesterday’s post asking why no one ever questions the fact that Amazon, a publicly traded company after all, never offers actual numbers to back up its claims, we offer a discussion of same from Michael Cader at Publishers Lunch. As he notes, there are “aspects of Amazon’s math that don’t add up, and probably never will given their limited disclosure.”

His analysis struck us as so brilliant it seemed to us to deserve to be quoted at length:

Checking the Numbers As Amazon Distracts with Press Releases

With Barnes & Noble expected to announce a color touchscreen device this afternoon, Amazon continues to push a steady stream of press releases to try to distract the media. Today they say they have completely redesigned the front end of their entire site to be optimized for iPad, though Amazon Windowshop app. Yesterday it was a Kindle classic, reaching for new ways of saying they are selling a lot while, true to tradition, not really saying anything.

Kindle svp Steve Kessel declares it “astonishing” that Kindle sales continue to grow, saying the new generation devices have “sold more since launch than we did during the entire fourth quarter of last year.” Well, the new devices have been out for 12 weeks, and a quarter is just thirteen weeks. And by exclaiming that the new-gen models have outsold Kindle devices from last year’s holiday quarter, it would seem that they are admitting that total Kindle sales declined in the meantime, in the two quarters after the holiday. Oops.

Other factoids are meant to underscore the growth of Kindle content sales. They say Kindle title sales for the first three quarters of the year have tripled compared to 2009. Among the top 1,000 titles on Amazon, Kindle sales have exceeded print sales. That’s helped considerably by strong Kindle demand for the very biggest titles–among the top 10 titles, Kindle versions are outselling print “at a rate of greater than 2 to 1.”

Does that make sense? Well let’s say you sell 50,000 hardcover copies of your new Big Author title. If you’re book is typical, about 35 percent of the total sale could be ebook, or another 25,000 units. On the hardcover side Amazon may be selling 15 percent of your opening units (7,500 copies); on the ebook side, they’re going to have 60 percent (15,000) to 75 percent of your sale (18,750). Voila, the 2 to 1 ratio.

Of course those heavily weighted top-of-the-list sales undercut the etailer’s other argument, that agency publishers are idiots for they way the price frontlist ebooks ….