BMW would kill to have 6% share of the auto market. When you're talking about a market as large as "automobiles" or "personal computers" that percentage represents a SIGNIFICANT number of potential customers. Too many to ignore when your product is already at least 80% capable of running on their hardware already. Also, the audience for an MMO is not "the internet." So those statistics are useless for the purpose at hand. You need to look at the specific data set, not use "google-fu" as you put it, and determine from there what kind of potential reach your product can have on the Mac. This real data is what the publishers go by - and that's why you see nearly every major triple-A MMO available for the Mac. And why Valve invested in bringing their engine to the Mac platform as well... too many examples to cite. MMOs are a unique market: The "typical gamer" does not live here. They are also consumed more as a hobby than a simple video game. In general, they also require more of an intellectual committment than a basic video game. They also typically have lower system requirements. They may also rely on players having more disposable income for an ongoing play experience (subscriptions or micro-transactions.) All of these are factors which make the target audience much more, well, targeted. And these factors also play well into the profile of a Mac user. I would say that WoW Insider survey isn't too far off the mark. I expect the Mac population of the WoW player base to be well above 20%, which shatters the notion of "typical" market share ideas. Specifically because of the factors above. No one is saying that the Mac is a bigger audience than Windows. But due to the size of the market, the numbers are not small. All major publishers (including NCsoft, by the way) have their major MMOs running on the Mac, and not because it's a bad business idea.