A Special Conversation: Fumita Ueda and Tetsuya Mizuguchi, Part 2

Continuing on from Part 1, the two master game developers, Ueda (Ico, Shadow of the Colossus) and Mizuguchi (Lumines, Rez, Child of Eden) move their discussion onto how they approach the difficult task of making their games. Let's listen in as they explain their creative process, shall we?

“I start with a brief image or picture, and then it spreads from there.” ~Fumita Ueda

Mizuguchi: Do you maybe feel more... full of life than when we met 10 years ago? Both of us, the width of what we're able to express in our games, and the types of messages we can convey... I think the creative work in game development has gradually become very fun and interesting. Of course, the actual work involved, the non-creative part, is still very difficult...

Ueda: Well, when we met, I think both 'Ico' and 'Rez' had a ….unique.... existence, even in the game industry, so.... I was always, always asked “Is this really a game?” back then, which I think might have effected my general outlook.

Mizuguchi: Ugh, yes. I had lots of people telling me “This king of thing isn't a game!” at that time. But right as that was happening, it [Rez] became popular on the internet. A lot of people started to say “This is a new type of game”, and they spread that impression to others. Also at that time, Youtube and other sites like it started, and everyone was able to convey lots of information using those new media outlets. The turn-around in opinion, that 'change in the wind' if you will, really made me want to try more and more new things. Since then, I've gradually become more capable in my job. Recently I've been getting really excited too, thinking things like “This 'making games' gig has really become interesting!”

Ueda: Recently, development tools and game engines, and people literate in making and using them, have been evolving and increasing. It's really convenient! Up until this point, we [Team Ico] had to first make the development tools, and then see what we could make creatively in the confines of the game engine's capability. Now, with how advanced and numerous development tools have become, we can focus more on the creative aspect of game design than ever before, without worrying about being restricted by the current technology. By the way, Mizuguchi-san, when you want to convey your idea, your image, of a new game to your staff, what do you do? For example, in 'Child of Eden' you have all these scenes with different colors and motifs, all represented abstractly, in quick succession. I'm interested in how you managed to convey all of that to your staff.

Mizuguchi: Well, when I first started development on 'Child of Eden', I sat down and wrote about 30 pages on the images I had floating around in my head. I then distributed that, in either English or Japanese, to every single member of my staff. I then told my 'Image Out' team I wanted them to help me visualize the game, based on what I had written. It ended up taking a lot of time; we were tweaking visual elements until the very end of development, but that's probably because I wanted the staff to use their imaginations to the fullest, so I actually wasn't that explicit about any details in my original document. In fact, I reduced the amount of information I provided, so that my staff could read between the lines and create something that exceeded my original vision. Despite that, though, I kept throwing out directions, like “let's try making it this way” or “let's try changing this” or “I want you to do that more like this”.... various things. That work requires a lot of patience and perseverance on all our parts to continue, it can go on forever.

Ueda: Did you use static images during the visualization process?

Mizuguchi: Yes, illustrations. One staff member drew over 1300 of them by himself.

Ueda: Whoa, in color?

Mizuguchi: Yeah, in color. He's still very young; he told me that when he was in high school, he played 'Rez' and decided that no matter what, he wanted to work with me. At first he had some difficulties, but he's not the type of person to give up easily. He just kept drawing, and eventually he not only put out visuals that exceeded my original image, he got so fast that he was able to make that many. For my staff, who works so hard and does things like that, I try very hard to stimulate their imaginations. Occasionally, we put out too many details on one thing, and then we expand on that image too much, so we have to tighten it up and shrink it down a little bit. That's what I've been doing with my team for the last 2 ½ years, basically. It's kind of like we made the game by untangling some sort of giant 'image puzzle' or something. Ueda-san, how do you go about creating your games?

Ueda: I always start out by drawing a picture. Then, I draw more pictures, continuing off that first one, until I have various scenes. Once I get a scenario that I like, and decide to make the game according to that series of images, I cut those specific images out and make a sort of 'trailer' picture for my staff, telling them “let's go with this kind of vibe”.

Mizuguchi: Do you have any hurdles when making the game meet the same feeling and meaning as the original picture you wanted to aim for?

Ueda: Sometimes, yes. I feel that rather than planning out every detail and scenario in the beginning, it's smoother to first create a picture representing what I want to express, and then letting the images swell from there, finally filling out and being reflected in the final game. It's also easier to add things later on this way, as I'll occasionally say “Hey, I want to draw this sort of scene or scenario” after we've already started production. My image isn't static, it changes and moves. I feel how I make games and how Mizuguchi-san makes games is actually pretty close. I mean, except for my initial drawings, which are pretty detailed, I don't give too much information out about how I want the game to go; my staff always manages to 'read between the lines' of my thoughts.

Mizuguchi: In game production, it's really difficult to be involved in all the various development stages at a level where you can give minute instructions for everything. Because of that, I think the best way to direct or produce is to gather a group of people who have the same vision for the game, the same way of thinking. For example, a group of 30 people, if they all have the same direction in their thinking, they can complete the game without blurring the initial vision.... whether that game will be a hit or not is another matter, though. Regardless, I think one of the most essential aspects of making a game is making a 'team'. Of course, my staff have their own respective and distinct personalities, but we're able to keep the big picture in focus, keep the initial vision intact, and then fine-tune it until the game is complete. I think this is really the core of game development.