GDC: Tsuchida, Shiraishi Talk Square Enix's New WiiWare Choices
The pair began with a trailer of Crystal Chronicles: My Life As a King, featuring the bright colors, petit characters and cheerful palette familiar to fans of Final Fantasy thematics. Unlike previous franchise titles, however, this one's bound for WiiWare, and producer Tsuchida opened the talk.
Tsuchida himself is a 20-year industry veteran, moving from working as a programmer on the first PC-Engine Final Fantasy, to, most recently, battle designer for the in-development FFXIII and producer on My Life As A King. The development process for WiiWare is different than any in his previous experience, he says.
First, however, he began with a discussion of Square Enix's methods and his perception of the company's strengths. "We are good at creating content with a large amount of high-quality CG," he said. "For WiiWare, you can easily download the game, which means that memory is limited. Which means you cannot use high-quality CG as a weapon."
"We couldn't put a lot of development effort to get ROI; we needed to create a game to create light users on the platform," he continued. Accustomed to big budgets and bigger platforms, he said, Square Enix's philosophy needed some evolution for the WiiWare platform. "But we wanted to maintain our strengths," he said. "That was a challenge for us."
In the standard Square Enix development process, he says, plot, characters and art are created first to set the world. Next come battle systems, field maps and cutscenes. The workflow is then created around technical restrictions, and from there the team develops graphical quality benchmarks. When the pipeline is established, the team mass-produces the content in a more efficient way.
200 people can engage in development of traditional Square Enix games -- 100 of them for over 12 months. "With this process we can only develop certain types of games, but if we are limited... we are probably not able to meet customers' needs, and the developer cannot fulfill his potential. I am always worried about it," he said.
With that in mind, says Tsuchida, the idea of retooling the standard process for WiiWare was appealing. Next, lead programmer Shiraishi took the podium. He was server and system programmer for Final Fantasy XI, having been with Square for nearly seven years. "During that time I built up quite a wish list," he said. "Being a server programmer, you really don't get to be part of the fun. I also wanted to try some new technologies that would be difficult in some of the larger teams. I also wanted to try making a game that did not rely on [data] volume... to think of ways to entertain with less data. As a result of these things, I wanted to try a smaller-size project with a smaller team."
That opportunity came in 2006. When Shiraishi first heard about Nintendo's WiiWare goals, he wrote a proposal the following day. Within days, he was at work on the Wii hardware. With no concrete information and prior to WiiWare's official announcement, Shiraishi says, "we had to assume a lot of things."
"In our current project, we knew we couldn't use our standard model. So we basically had to go back to the basics." First, he says, they decided on a target price and memory size, and then went back to some older assets for ideas. Once a basic pipeline was established, new assets could be added.
Development was scheduled for seven months, with a maximum of eight people. This time, the team started with the game concept rather than with visual assets. It's about building a world rather than exploring one, so the need for extensive concept art and new assets was minimized. "Unless you can show something that's truly compelling, we think it's better to leave things to the imagination," Shiraishi suggested.
Because the game design was somewhat new to the team, they chose Squirrel, a lightweight programming language that would allow for plenty of last-minute changes."We asked [the designers] to limit their scripting to the fairly minimal," Shiraishi explains. "At Square Enix, we don't have a culture of using other people's codes. But I thought Nintendo Ware would be a good fit. I think gone are the days when developing everything in house is a good idea."
Division of labor lines were also blurred -- the game designer also created menus and visual effects, and everyone contributed to the design.
Next, Shiraishi showed a demonstration of the gameplay, in which a diminutive king must handle affairs within his small, walled city, understanding the needs of the heroes and deciding what to build and invest in. It's a colorful "kingdom simulation" that still manages to retain the traditional Square Enix aesthetic.
So the team did struggle with narrowing the target audience, keeping cost low and meeting a quicker and lighter schedule -- for example, they spent months on a battle system only to scrap it entirely, and development took about a year, not the original seven months scheduled. Nonetheless, it looks as if the developer of traditionally gigantic and CG-heavy projects was able to successfully retool its design process specifically for WiiWare, and learned many things in the process.
"I didn't plan to make a 'Square Enix game'," Shiraishi concluded. "But we ended up with a game that's unmistakably Square Enix."