While writing the most recent Shifting Perspectives column and browsing old records on Druid population statistics, I started to wonder about the various factors that play a role in how popular a class becomes. While Blizzard and Blizzard alone has the exact numbers on who's playing what, various fan sites have honed data collection strategies over the years and amassed a pretty impressive pile of numbers. This only got easier when the Armory launched in spring 2007, and by now I'd be surprised if players weren't at least broadly accurate about overall trends. If we can trust what we see, how we do best explain fluctuations in class popularity? Has Arena success (or the lack thereof) been as influential as we think? Is class population an accurate, albeit crude, guide to the overall "quality" of a class at any given moment -- or just a guide to the perceived "quality?" I'd be interested to hear what people think.
Having played a Druid since the beginning of Burning Crusade and observed it going from the second least-played class at 60 to the third or fourth most-played class at 80, I have my own theories about what's influenced Druid population numbers particularly, but I need to do a little more digging before I can be sure. However, I don't know whether any of it really applies to other classes, and the meteoric rise of the Death Knight is a thought-provoking (and somewhat troubling) trend.