StarCraft II could have been a prettier game, with massive Zerg Ultralisks and screen-filling Protoss Motherships. It could have been stuffed with new units like the Cybercat and Hercules transport ship. Blame designer Dustin Browder and eSports for making StarCraft II what it is: clear, simple, uncertain and a game that demands skillful players.
Browder said at his GDC presentation that trying to design StarCraft II's multiplayer side was "insanely hard" and "like inventing Basketball 2." But in 2005, he and Blizzard Entertainment believed that "maybe, if we really hoofed it, we could ship it in 2008." Obviously that didn't happen.
When the StarCraft II team began its work on the game, it looked at its competition, namely Relic Entertainment's Warhammer 40,000: Dawn of War and Gas Powered's Games' Supreme Commander. Dawn of War featured four playable races to StarCraft's three. Supreme Commander drowned the player with playable units, with more than 150 at a player's disposal.
StarCraft II would stick with three races—Zerg, Protoss and Terran—and a total of 45 units. Less perceived content than its competitors, Browder said, because StarCraft II needed to be a sport, watchable and entertaining enough that people would travel for hours to watch other people play.
To make it watchable, it needed to be visually clear. "This is why the artists hate me," he said. Browder showed concept art of the 300 ton Ultralisk, with a tiny Protoss Zealot to illustrate scale. Then he showed the in-game Ultralisk, tiny and unimposing. Browder explained that having an Ultralisk to scale would never work. Players could hide 20 Zerglings behind a beast that size, sowing visual confusion.
Units in StarCraft II, Browder said, needed to be big, but only "as big as we dare."
To make StarCraft II a successful sport, it also needed to be simple, with just 12 to 15 unit types per race. "This is why the designers hate me," Browder said before highlighting all the unit types that Blizzard cut from the game's multiplayer component. Some of those units wound up in the game's single-player campaign, but Browder and his team wanted to focus on a small collection of units.
"Too many units causes confusion," he argued, likening StarCraft II multiplayer to another sport: football. "In football, there's one unit: the human. Of course, there are wide receivers, quarterbacks and other roles, but when you see those game pieces moving around in a football game, you know what they're capable of."
He said that a small selection of units still left the game open to complexity, explaining the relationship between Terran Marines and Zerg Banelings. Banelings, with their explosive acid attacks, are a great counter to Marines, melting squads easily. That is, unless the Marines have the Stim Pack upgrade that allows them to do more damage and move more quickly. Then they're the dominant unit. That is, unless players invest in the Banelings movement upgrade and Creep bonuses.
Those factors lead to uncertainty, a key component in keeping StarCraft's eSport qualities intact. Skill through micromanagement and macromanagement, flanking and terrain tactics, skills other real-time strategy games tried to "scrub out," he said, was also crucial to StarCraft II's maintaining its sports-like nature.
Uncertainty is also one of the main reasons the dreaded Zergling Rush—a sprint to build a small army of basic Zerg units and attack—remains in StarCraft II. Browder said the Zergling Rush adds uncertainty from the get-go. "By delaying the rush, you're just delaying when the fun begins," he said.
Blizzard's drive to keep StarCraft's eSports nature intact for the sequel affected every component of the game, Browder said, from UI to story to mission design. "Was it worth it?" Browder said. "It totally was. This stuff is cool."