I see. And so, you go through the concept phase, and then the phase you're in now -- would you call it prototyping, or would you call it...

JC: Well we've passed our prototyping. The concept started more by myself. Because at that time, people were still finishing up flOw; I am the only guy who has time. So I basically prototyped myself, and came up with a series of prototypes, and said, "Hey, this is what I envision, and these are the key technologies that I hope we can have."

Then, once the team started, the whole would basically divide the production into the prototype phase, and then the production phase. We finished our prototype phase in six months. It was a long prototype phase, because we originally thought the production would not be that long; probably another six months, and then we could be done. We hoped that we could finish the game in a year.

But after the prototype phase, we realized that there are still a lot of unknown things, because this game is just so different. But we've started production now, and we've started making levels, but as we are making levels and doing play tests, we notice more issues coming up, and to solve them, it requires more prototyping.

So it's more like an iterative process rather than, "Hey, we design, and finish this, and then we execute." I think it's just not going to happen to any innovative games. You know, if it's something new, stuff just changes all the time.

Do you find your process is still somewhat getting figured out, since you're a newer company?

JC: Well we've started this process since we were working on a school project -- we had the [first] game, the Cloud game... You know, it's like, at the beginning, we start, we have enough prototypes, like 10, but then as we start production, it's just taking longer than we thought, and then it requires more prototypes.

It's basically like what we were taught in school: design, production, and play test; redesign, production, and play test; redesign, production, and play test. It's an iterative process. At the beginning it's really big, it's a very vague design, but then it's like a pyramid, and in the end it's a really small iteration.

Refined.

JC: Yeah. Which is not really like the process in big studios and big projects. We kind of -- both Sony and us -- we are all learning, because through flOw, we realized that it's not going to be a very straightforward process. So, I think right now, even though we are in production, there are still a lot of things coming up. But everybody understands it. So it's cool that way.

thatgamecompany's flOw

Is it at all frustrating to be making the same misjudgment each time? Like when you said you think, "Okay, it's going to take this long," but always it takes way longer.

JC: This time we actually learned that, because during development, people crunched so much. Even though I was not there most of the time, I can feel it. We originally predicted that we could finish flOw in six months -- in four months, actually. We started production in June or July, 2006, and at that time the PlayStation was launching at the end of November. So we were like, "We're gonna have to catch up so we can be a launch title!" So that's four months, right?

And because nobody has worked on PS3, well, it took me two months with Nick -- we both had no experience in Flash at all; we learned Flash and made the game. And so, because we were also in school, we figured if we really calculate all the time we spent, it's probably three weeks. So, "OK! Definitely we can get there! Now we have five people! We have four months! Sure! The game is already done, and everything's all there!"

But in the end, the game took us eight months to make, and we only delivered half of the original design. So it's like we 400% under-delivered. So you can imagine how much crunch the people had. The launch date had been delayed several times; each time it's a big morale hit.

So from that experience, the team members... Me? I haven't been through that hell, so the team members will always be more conservative than me. And I think because they have a more conservative attitude this time, we didn't crunch at all. We always delivered in time. It's like 40 hours a week -- you know, 50 a week for some programmers, but we're proud that so far we didn't crunch at all.

That's good.

JC: But, you know, the games definitely took longer to make, but we are not going to kill anybody.