Many people have tried to identify the secret of Blizzard's success, a topic guaranteed to interest a lot of game developers. Still, when someone who has been around as long as Gordon Walton gives it a shot, it's worth paying attention. Not only is he a funny man, he knows what he is talking about.

Worlds In Motion has an article about Gordon's talk about World of Warcraft at the Austin GDC a couple of months ago. Here is my summary:

1. Learn from the past. Anyone remember the idea of World of Warcraft being against conventional wisdom when it was first announced? Surely Blizzard would fall on their faces making a fantasy MMO...
2. Aim for a broad user base.
3. Quality counts. As someone once pointed out: World of Warcraft was the first MMO developed by a company that knew how to make AAA offline games. I have been playing a lot of virtual worlds and Asian MMOs recently, and boy, they suck. In terms of user interface design and player guidance they are miles behind the top offline games.
4. Support solo play. I wonder if this was an innovation because MMOs before WoW tended to come from MUDs?
5. Simplify the GUI. If you've played WoW for a while, you might know how odd it feels to go from a level 70 raiding character to a lvl 1 nub with no add-ons. You suddenly have only 3 buttons. I really like Gordon's remark about the GUI customization being a release valve for the hard core. It's also great for Blizzard: they've taken most of the really popular add-ons and put them into the standard GUI.
6. Tons of content. WoW was way more expensive than I expected. I have heard numbers between 50 and 80 million dollars. One thing that pays for is content.
7. PvP content. Blizzard had lots of experience here. It's easy to underestimate everything they learned from Battle.net.
8. Don't tune for the hardcore. Yeah... I know people who still haven't gotten to level 60.
9. Let them quit. Gordon kinda makes the same point in this and the previous lesson: You can force people to do stuff by clever game design, but they will probably hate you for it in long term (and in fact complaints about grinding are the number one reason I have seen for why people stop).
10. Offer the right amount, and the right kind of choices. This is worth a couple of blog post of it's own.
11. Easy to learn, difficult to master. More on this below. I am still learning new things about WoW after almost 3 years. Like why parry mechanics make it a bad idea to attack mobs from the front when you're not the tank. Or how you can increase your DPS by 5-10% if you understand how the client communicates with the server.
12. Brands matter. This is the most secret ingredient of Blizzard's secret sauce: how they got where they are.

What I consider one of the most impressive parts of WoW is the progression design. How you're taught about the game. How you're led through it. How there are always just the right number of carrots dangling in front of you. This comes up in lessons 5, 6, 10 and 11. I think the progression design in WoW is perhaps the best I have ever seen in a game, even beating Zelda: A Link To The Past. (I've been meaning to analyze how both games do this - now there's some useful games research.)

The article concludes with:

The thread that tied the talk together was changing the mindset of the developers: it's about understanding that a general audience is not the power gamers. If a game is to be successful with a broader audience, it has to be more fun, more directed, more accessible, and faster-paced.

Emphasis mine. Think of lesson 8: Don't tune for the hardcore. But hey, developers are the hardcore. And the hardcore are the first to set up camp in your forums and tell you what they like and don't like about what you say you are building. (This was first pointed out sixteen years ago, by Chris Crawford.) It takes a clear head and nerves of steel to ignore all that. Blizzard uses mantras for design: that's a great way to continuously remind yourself what you set out to do.