How WiiWare challenged Square Enix to rethink teams

How WiiWare challenged Square Enix to rethink teams

By Ed Fear in San Francisco

February 22nd 2008 at 3:05AM

GDC08:My Life as a King's effect on the Japanese giant's development process

In a pre-postmortem of their upcoming WiiWare title My Life as a King at GDC 2008, Square Enix staff members Toshiro Tsuchida and Fumiaki Shiraishi have discussed how the prospect of making a downloadable title for WiiWare caused a re-evaluation of team structures at the company.

The idea of creating a WiiWare title came to Shiraishi, the game's lead programmer, before the service had ever been announced. After reading on the internet that Nintendo would be embracing the digital download model in 2006 - the year before the service was officially announced - Shiraishi wrote a project proposal within the day and showed it to producer Tsuchida the day after.

Shiraishi, who has previously worked as a server programmer for Final Fantasy XI and Front Mission Online, had become increasingly disillusioned with his position and the lack of input he had to any project. As such, the idea of having a much smaller team working on a game interested him greatly, and he built the proposal estimating the development time at seven months and only requiring a single designer, two programmers and up to five artists.

The traditional team structure at Square Enix dictates strict roles, where between 100 and 200 staff members work for several years on very narrowly defined roles so that they can hone their craft. How Square Enix differentiates itself, Tsuchida explained, was that it prided itself on being able to make large quantities of high-quality assets, which empowered them to be able to build new worlds.

"Square Enix had been following the same development model since Final Fantasy 7," he continued, "but the tiny size of a WiiWare download meant that we simply couldn't rely on large amounts of CG as we had done."

As such, the project was greenlit with the premise that the staff minimise the risk involved as much as possible, as the company was unsure of such a radical departure from its normal working practices. Rather than creating a large amount of assets at the beginning of the project, as most of the firm's other titles do, Tsuchida and Shiraishi managed to get permission to use existing Gamecube assets from Final Fantasy Crystal Chronicles to use for the title.

Similarly, the team deliberately narrowed down the game scope to keep it manageable with such low staff members. Rather than focusing on adventuring, the team decided to make a game about building a town. Although initially there was a battle system and a world outside of the town, these were eventually scrapped to keep the project manageable. As Shiraishi explained, though, they didn't view this as a bad thing: "If you can't show something really compelling, I think it's probably best to leave it to the imagination," he said.

In addition to scaling back, risk was reduced by blurring the roles of staff to keep the project size low, meaning that many people had several roles on the game. The team also used NintendoWare to speed up development, as well as coding most of the title in the high-level scripting language Squirrel, and they also attempted to use procedural content as much as possible.

In the end, the game took 17 months and 18 people - more than the estimate, but significantly smaller and quicker than traditional Square Enix projects. Enumerating the difficulties in the development process, Shiraishi said that although starting a project before the download service had even been discussed was risky and, while it paid off by letting Square Enix get a title on the service near to launch - something the company is never normally able to do - not knowing what the WiiWare audience would end up being like and not knowing how the competition would look caused the team to re-do a significant amount of work as it kept changing its audience and its expectations of the competition.

Nevertheless, Square Enix is happy with how the project has progressed, and shown those skeptical in the company that not all projects need to be monolithic. But most of all, Tsuchida concluded, it empowered its staff in a way that they hadn't seen when they were all small cogs in the machine.

"While specialisation helps people hone their high-end skills, by making everyone use a wide range of skills every person had a say in the project, which makes them happy and gives them confidence. Having a good opportunity to make use of their skills makes them happy, and this will inevitably be shown in the product, which will in turn improve their evaluations and make them even happier."