GamesIndustry's Rob Fahey penned an editorial on World of Warcraft's amazing success, inviting industry insiders and writers alike to not take it as an anomaly, impossible to explain, but rather to accept that its commercial success points to the fact that some of our accepted wisdom is actually not quite correct. Here's a snip:
Like many fields - most fields, in fact - we're not good at slaughtering our sacred cows. WoW's success defies conventional wisdom, which means, in the most honest and direct terms, that the conventional wisdom is wrong and must be re-evaluated - but it's far easier to shrug, brand the game an outlier, pay lip-service to Blizzard's "unique" talent and be done with it. In a sense, I think that does a great disservice to Blizzard; it pretends that their success is down to some kind of magic dust sprinkled all over Azeroth, attributing the success of the game entirely to the spark of genius in its creators' minds. Nobody is denying that spark of genius; but the 1% Inspiration, 99% Perspiration rule still applies, and to deny the 99% of Blizzard's efforts which are down to bloody hard work and an extraordinary process of learning and refinement seems deeply misguided.
Let's frame this in more concrete terms. What is it that WoW is doing that's so far outside the beliefs of so many people in the games business? Perhaps the most obvious and relevant example of that is that, eight years after launch, the changes being made to World of Warcraft are still incredibly fundamental (a complete repudiation of the idea that if it ain't broke, you don't fix it) and moreover, continue to be aimed at making the game more accessible and more appealing to casual audiences.
The first of those things is a really interesting and rare interpretation of the concepts of data-driven design. Blizzard probably has more data about what players actually do in their game than any other developer on earth (with the possible, but arguable, exception of Zynga). As developers have come to grips with the idea of using that data to drive design, the general idea I've heard expressed is one of continual and gradual refinement - the concept being that there's one "best" way of doing things, and you're using the data fed back from your players to gradually hone and refine the game so that it gets closer and closer to that ideal system.
Blizzard uses data in a much more aggressive way. Mists of Pandaria rips apart underlying systems which have been in place for years; the previous expansion, Cataclysm, also made enormous changes. After eight years, you might expect that a process of data-based refinement would have narrowed down to tiny tweaks - as indeed has been the case in the majority of other online games that have run for more than a few years. Instead, Blizzard uses that data to suggest and implement entirely new approaches, allowing it to keep the game experience fresh and challenging - preventing single, over-optimised approaches to play from emerging and players from getting bored. Moreover, the company understands that the arguably risky nature of extensively changing game systems is actually mitigated by its ability to capture and understand data, because any major errors can be spotted and dealt with quickly.
There's a lot more over at GamesIndustry, so I invite you all to read the actual editorial and share your thoughts on it.