When Mark Twain popularized the phrase "lies, damn lies, and statistics", he was referring to the way that one can selectively report statistics and tell just about any narrative desired.
We see this everyday in the mainstream media when they report on AAP revenue stats (to name just one example), and even in the trade press. Remember when Jim Milliot claimed that Amazon would soon be the fifth-largest bookseller in the US? Jim selectively reported publicly available stats in order to tell the story he wanted (one that wasn't even close to being true).
Statistics are used all the time to mislead, but some stats don't need any help.
There are statistics that mislead you, confuse the issue, or distract attention all on their own. For example, every month Publishers Weekly reports the US Census Bureau's estimated bookstore revenues. You might think those stats reflect book sales, except that the estimates include more than just book sales (and they are about a third as large as they were a decade ago).
Statistics can mislead all on their own, if you're not careful, and that includes stats frequently thrown around in debates about the book publishing industry.
For example, just the other day a reader made an off hand mention of the stat:
Print is still 75% of all publishing sales, including indie sales. It’s holding steady, and may be increasing.
While this stat is mostly true (let's not nitpick the specific value) it is still misleading and distracting.
This statistic misleads you into only looking at book sales, and it distracts you from seeing industry trends. For example, textbook sales aren't going digital because they are evaporating, and being replaced by free OER downloads. Nonfiction print sales are being cannibalized not by ebooks but by freely available info on the internet as well as videos and webinars.
This stat also misleads you into thinking that it applies equally well in all categories and genres, when each category is unique, and some have actually gone overwhelmingly digital.
This stat is a doubly whammy; it keeps you from seeing both the big picture and the specific details. That's why one should always take care when using it.
That rule applies to all statistics, really, and it's worth keeping in mind when numbers start flying around.