This was on the Passive Voice yesterday--the original article is here, and it's a survey of e-book buyers by Bowker Market Research to find out what kind of device they use.

Barnes & Noble’s Nook devices had a 14% share in the second quarter [of 2012], a figure that has held steady since the fourth quarter of 2011, but was down from a peak of 22% in the third quarter of 2010

Now this is a different number from either number that appears here: This survey includes devices like desktop computers, smart phones, and iPhones. But 14% is a long way from 27%--a number that is supposed to also indicate the percentage of the e-book market controlled by Barnes & Noble, even though nobody knows how big that market actually is.

If the recent presidential election taught us anything, it's that my jokes have a strange way of coming true--I mean, it's that decisions made about what to measure can drastically affect results. So, deciding that a smart phone or a desktop computer is or is not an e-reading device, or deciding that e-books sold by Amazon don't count--how one chooses to define the "e-reading market" or the "e-book market," in other words--can drastically affect perceptions of how a company is doing. And just like there are various pollsters who make money telling presidential candidates exactly what they want to hear, companies that report data are serving a market that rarely consists of disinterested observers.