I'm skipping this month's report, because the numbers seem to be out of whack. (By way of example, usage of OS X 10.9 declined significantly, and OS X 10.1, which has never been on the list before other than in "Other", showed greater usage than OS X 10.4.)
I've always wondered how reliable the numbers I've been posting have been, but last month's reported numbers have really got me scratching my head.