It is interesting to see the evolution of these games. I saw a quasi tech side FarmVille postmortem and they really were bolting things on to it as it got more popular. "Oops! We gotta do some more stuff here."

RG: And it looks, and feels, bolted on.

Right, so that accounts for the strange UI. But it is interesting to see this as it starts to be more and more from a web development side, how the UI improves, because people have appreciated now that simple is better, and have figured out how to actually communicate this information a little better. I think that's the hardest thing to do.

RG: Well, and I think another big lesson for all games from social media is the constant teaching of the UI.

For example, a drag and slide bar pretty much doesn't exist in social media games. There's a right and left arrow to go to the next page, but there's not like a slide bar to go through a long piece of information, because what people have found is a lot of people just didn't figure it out, you know what I mean?

It's a level of user interface sophistication where they lose people. When people have done that, people buy all the stuff that's on the first page but they never buy the stuff on the second or third page, because they never really drag that slider bar.

And so, making sure your UI is really easy to understand, we're talking about it has to be, you know, when you walk up to a washing machine or a dishwasher, everybody knows how to operate it, because it's got a very simple set of standardized controls. And there's no problem with that, it's actually just a reality! If you're going to deal with all of humanity, just make sure that you hand-hold people through the utilization, and make things as standardized, and only one level of conceptual complexity, as possible.

It doesn't mean they don't appreciate good literary content, it doesn't mean they don't appreciate rich, detailed NPCs in immersive virtual worlds -- you just have to don't overwhelm them with it. Even as a hardcore gamer, I play some things -- like I download many of the free-to-play MMOs that are out these days. A lot of them are graphically beautiful, have really nice user interfaces, have NPCs of great diversity, have all the kind of what I call "feature complete" in the sense of avatars with detail, and just all these other kinds of ways that they're rich.

But I find them overwhelming -- you get dropped in the middle of this gigantic world, you have no idea what to do, you have no idea where to start, you're going like, "Oh my God, it's going to take me so long to even know whether I like this game, that I'm daunted to start!", if you know what I mean. I go, "Yep, looks beautiful! Okay, I'm done." [laughs] And I don't play 'em. Even the kinds of games that you might think I would make, I don't generally play, because they're often just too much of a hassle to get into them. And so comfortably introducing people into your depth is a lesson all games should learn.

Yeah, and they have the advantage there of metrics. Being able to see what people are actually doing, they're not just guessing...

RG: Right, and by the way we had that in MMOs also. So no MMO developer should be excused from not knowing the answers to these questions because we had metrics then, too.

I think back then there was more of a reluctance to do it because it felt like cheating or something, as a game developer. But it is quite useful.

RG: Yeah, well, I can tell you in Ultima Online we constantly used those metrics to redesign the game. For example, one of my favorite stories is, in Ultima Online, when the game shipped, you could use a fishing pole on the water and there was a 50/50 chance you'd get a fish. Beginning and end of simulation -- literately use a pole, on water, 50/50, fish. Lots of people did it, tons of people did it.

And people began to believe apocryphal information about fishing; they began to believe that if you fished in a river versus in the ocean they were better chances of getting fish, which of course was not true. I told you the simulation use fishing pole, on water, 50/50, fish. That's it!

But so many people were doing it, and so many people had these fictitious beliefs that we thought, "Wow, we should spend some time to make fishing better!" And we did. Over time we actually made the fishing simulation more improved, gave you different kinds of fish, and there really was a point to using different places, and then it became even more popular.

And there were things that we thought were really cool that we put in the game, that nobody noticed or cared about -- very sad and tragic. But we either fixed and adressed those, or often, we just removed them from the game.

And so, even from the beginning -- I don't think of the metrics-driven feedback to me as new at all; I actually look at it as, "Yeah, of course you're doing it." And they are doing some things that are new, like we would never do A/B testing on the color of text. But you know that's interesting and useful to know how to optimize your value stream, but it really is something we were already doing 10 years ago.