Time is money.
For game developers, the typical time frame for making a game is 18 months, a good chunk of it spent on programming the game's "engine," the technical foundation that must be built before even the first polygon is drawn.
With a modest team of 25, that amounts to about $2.5 million in development costs. For each month that the game is delayed, add $200,000 to the tab. No wonder the final hours of game production are often marked by all-nighters, caffeine splurges, cold pizza and neglected personal hygiene.
One advantage of the next-generation consoles--GameCube and Xbox in particular--is that they help with the technical heavy lifting, so developers can get on with making the game. This was the case with "Star Wars Rogue Leader: Rogue Squadron II," made by Factor 5 and published by LucasArts Entertainment Co. for Nintendo's GameCube. From writing the first line of code to the last day of testing, the game took about half the time to make as the original "Rogue Squadron." Also developed by Factor 5, the original took 18 months.
Julian Eggebrecht, president and co-founder of Factor 5 in San Rafael, Calif., talks about what it's like to make games for the new, ultra-powerful consoles.
Question: How were you able to cut the production schedule in half?
Answer: The first "Rogue" took 18 months; 12 months of that was engine work. So for a year, there was nothing on the screen. On the GameCube, you can cut that down to one month. What you had to program yourself is now in the hardware, that basic programming layer. Nintendo optimized that with the GameCube, streamlining that very nasty step that takes forever. Once you have an engine, everything else goes much faster. This is true for both the GameCube and the Xbox. They cut out that initial research and development time, and that was crucial.
Q: Does this mean your costs are lower?
A: Not necessarily. You've got to have more artists now. What took us forever was building the ships in the game. That was by far the biggest amount of work. In that sense, this generation of games is 100 times more work-intensive simply because you can do so much more. It's a lot more work to create the artwork. The art process was the biggest contributor to the development costs now. "Rogue Leader" cost the same in nine months as "Rogue," which took 18 months.
Q: How important is the controller in designing games?
A: It's crucial. The controller is the interface between the gamer and the game. The GameCube controller minimizes the number of buttons to the ones that are absolutely necessary. First, you have a huge A button, which is your main action. Then if you need others, you simply slide your thumb over to the right. It's very intuitive. On the Xbox controller, it's hard to distinguish which button is the most important one. They're all the same, and they're all very small.
Q: How was it to work with a license that's as well known as "Star Wars?"
A: It's the greatest blessing and the biggest nightmare. As long as the movies play along, that's great. But if you hit something in the movie that doesn't fit your game, you have to weasel your way out of it. In the Hoth battle, for example, you have to lose. You cannot win. It would be against the "Star Wars" canon if you win. But the game has to give the player a good feeling even though they've been swept off the planet. So the mission is to get off the planet and save as many people and transport vehicles as possible. The player feels, "At least I made it out alive and I saved transports." It is an ambiguous feeling. But then again, "The Empire Strikes Back" is mostly a dark movie. Most video game stories are so plain that it's actually fun to do something negative.
Q: What was the biggest technical challenge you encountered in pulling the game together?
A: The biggest problem was optimizing the CPU power. In one level, we had 300 troopers running around and two wingmen, each operating with their own artificial intelligence. We also had 500 lasers firing off on the screen at any one time. The sheer number of elements controlled by the CPU is tremendous. On top of that, we had an evacuation scene running in the background in real time. The other challenge was the graphics. Because the machine is so powerful, it was mostly an issue of having enough time. We needed to pull off something quickly. We needed a shortcut. So we used one of the GameCube's strengths, being able to slap on several textures on top of each other. After a few layers, it starts to look real. But really it's just a canvas of four or six individual textures. That's it.
Alex Pham covers the video game industry. She can be reached at firstname.lastname@example.org.