The Issues of Games
An interesting early-morning panel today at Siggraph was the videogame panel Games: Evolving on an Order of Magnitude, which featured apropos comments about where the industry is going technically and from a business point of view from some heavy hitters in the game world. The panel, moderated by Michel Kripalani of Autodesk, included Lyle Hall of THQ Inc., Martin Walker of Artificial Mind & Movement, Steve Theodore of Bungie, Steve Sullivan from Lucas Arts, and Jeff Lander of Electronic Arts.
What I found most fastcinating was actually the business discussion. The panelists, and many in the audience asking questions, brought up issues revolving around topics of workflow, scheduling, R&D, budgeting, and so on–issues they all claimed are becoming increasingly complicated for videogame companies. When to spend time and money on automated systems versus investing in more manpower? How to balance generally process over automated technologies? How to split resources between finishing major projects on tight deadlines and tool R&D work for the next project looming on the horizon? How to manage the schedule and make sure your artists don’t get burned out?
These questions and the panel’s answers illustrate that videogames are, indeed, a business. A major business, doing more box office, so to speak, than all the world’s feature films combined. And yet, to hear the panelists talk, it sort of seems like the creative and technical innovations coming out of that industry are far ahead of the logistical business maturation the industry now needs to manage people and pipelines that are now working around the clock on these kinds of projects.
For example, Jeff Lander talked about the need to combine automated processes, like monitoring nightly level builds, with the human element–having what he called “an imbedded tester” built into your infrastructure to catch the kind of nuanced problems that automated systems can’t.
“There are some things that computers just can’t tell,” he related. “It can tell that, sure it runs and the character walks, but it walks with its foot over its head. But then, we need someone from the team to come in–ideally, we like to have an imbedded tester from Day 1, which is really hard to do sometimes. But, you have somebody sweep some portion of the game, and then validate the nightly build as good, and then it gets promoted to a ‘known good.’ ”
And Steve Theodore added, culturally, a more permanent and ingrained testing mindset needs to be incorporated into people’s work procedures.
“One of the things we learned from our association with Microsoft is that the biggest mistake you can possibly make on this kind of project is to think that testing is something that happens at the end of the project. Testing is always going on from the beginning to the end. Every aspect goes through a carefully done test process. That is amazing to me, and so valuable, and yet not everyone does it.”
And, adds Lander, “This is something you want to do even if you are an indie studio–three guys in a room.”
And yet, as they discussed over the course of the panel, budgets and timelines for such projects automatically lead people running those projects to desperately need automated solutions whenever possible. In other words, it’s a dichotomy. Sounds a lot like the feature film world I’m a lot more used to covering.
–MG
