Do you think that you're evolving quickly because you're shipping quickly? Is the gameplay evolving at the pace of the mobile market?

DM: I don't know. That's a good question. It might be too early to tell. Shipping a game in six months is definitely not my favorite thing. We've definitely done it, and we've now done it twice in a row, and there is something to it. I don't know about this six month thing, but there's definitely something to saying, "We're going to be in production in this game for this long, but we're going to ship it here." [Mustard moves one hand to show the full length of production and then uses the other to mark the mid-point in the cycle.]

We're going to ship it before we've even done our development phase. I think there's something really key. I think that a lot of game developers are starting to think more that way, where shipping the game and finishing development on the game don't happen at the same time.

It's like you're winning your first battle, but the war continues.

DM: Right, the war continues. Because that really does allow you to find the fun of the game. You make sure there's enough meatiness there. That's the way that I think game development needs to move forward, in order to really make more refined games that meet the needs of the audience better.

This isn't like film or books, or set media. I think we need to allow our communities to be more involved in shaping the experience that they want, because it is interactive, right? And the best way to find out what your game needs is to get millions of people using it.

So in that respect it does allow it to accelerate much, much faster. So many of the decisions that went into Infinity Blade II were based on those several months of making DLC content for the first game, because we could react so well. It's why Valve's games are so polished.

The kind of metrics and the data they're able to get from Steam really allows them to look at how people are playing their games. And we were able to do some very similar stuff with the first Infinity Blade, where we got it out there, and we were getting all sorts of data back on where people were doing things, where people were getting stuck, or not stuck.

To the point where Infinity Blade I is now a refined enough experience that over Thanksgiving Geremy gave an iPad 2 to my dad. Now my dad, he probably hasn't played a videogame since Ms. Pac-Man. He's late fifties, hasn't played a game ever.

I handed him an iPad to play Infinity Blade. When we were balancing we're like, "We want a non-gamer who's never played a game to be able to be good enough that on the fourth Bloodline, they kill the God King." To us, that'd be the perfectly-paced experience, because that will give him enough failures at the God King to really feel like it was a payoff when they beat him. And my dad, who, again, hasn't played a game probably in 20 years, beat the God King on the fourth Bloodline.

And I'm like, "Oh my gosh! We did it!" We made it. And I doubt he would have been able to do that on the Infinity Blade that launched on December 9th last year. Maybe, but we were able to balance enough stuff, and tweak enough things, that it became this more refined experience. And so I think that's hugely valuable.

I think in retrospect, having done it twice, that our development cycles are a little too short. Not that the games are less polished because of it, but we're way more burnt out. Because, in order to make II feel the experience it needed to be, it required way more crunching than is effective. It required for us, for the last two or three months, to just death march kill ourselves. I mean guys are just working so many hours, doing so much, and that's not really good, I think, for the longevity of our studio.

I don't think it's Epic's style, either.

DM: It's not Epic's style. It's not, no. We don't look at that like that's a good thing at all. We only did it because we definitely, passionately wanted to get the game done, and we wanted a little more in there... It happens when it's that short of a development cycle. Stuff happens so fast.

And so we definitely won't do that again. It's not worth the cost. I would rather take an extra two or three months than burn the guys out, or burn even me out. It doesn't allow enough time to sit there and let the game breathe.

When using metrics and player feedback, are you looking for consensus? Are you looking for things that please the most people, or are you looking for the best ideas that come out of the noise?

DM: We have a lot of confidence in our ideas at Chair. So it's definitely not about looking for consensus. It's much more about trying to observe, as best as we can, how people are playing the game, and finding ways to better shape their experience.

There are a few things like "I wish my dodge buttons could be bigger," that we're like, "Oh, yeah. Okay, that makes sense to us. Sure." But that's like the edge case. There's way less of that, and more of us just watching and going, "Oh, you know what? People would have more fun if we gave them a little more gold right there," or "You know what? That weapon, based on how people are playing it, it's actually doing too much damage." It's actually detracting from their fun, because it's too easy right there, that it's actually less fun than it should be.

So, it's more just we've put in a lot of metrics to allow us to observe how people are playing, and tweak it based on that. There's much, much more of that than "I wish that this was blue instead of red."