Imagine a real-time strategy game that took place in full three-dimensional space, where your units weren't just AI-controlled automatons, but real people. Those people were playing a beautifully rendered space combat simulator. You could give them orders, but their ability to carry them out was based on their skill as a pilot. Unlike in StarCraft or Command & Conquer where all units of the same type were created equal, here the best dogfighters could decimate half of the other team before being banished to an escape pod. Such a game came out ten years ago, and its name was Allegiance.

The game had everything - a fully realized RTS game, complete with tech trees and resource gathering, and a space combat simulator with (for the most part) accurate zero-G physics. But less than a year and a half after the game's launch, the number of active players had dropped to the point where Allegiance's publisher had dropped support, and the community had dwindled to almost nothing. What went wrong?

The origin of Allegiance is one of the more fascinating tales of game development; it wasn't developed by a gaming studio powerhouse, or by a big name in game design. Instead, it was developed by a group that had never shipped an actual product, let alone a videogame.

Microsoft Research was the skunkworks of the software giant, a place where cool and forward-looking technologies and concepts were developed to later be rolled into Microsoft products. This was the team that developed such technologies as Digital Ink for Tablet PC and Microsoft ClearType. What were they doing making a videogame?

"It turns out that computer games usually lead other product categories in using cutting edge technologies," says Joel "Solap" Dehlin, executive producer on Allegiance. One of the VPs of MSR, Rick Rashid, who had developed a multiplayer networked game for the Xerox Alto called Alto Trek, had been keeping the code alive in his own time for close to twenty years, and thought it would make a cool project for MSR. Rashid, along with Dehlin and Rob Girling (who would become Lead Game Designer), hashed out their ideas. "Rob and I would spend hours talking through ideas. We'd go out to a pier on Lake Washington or over to my house or wherever. Then we'd come back and talk to David [Pugh, one of the head programmers], who would tell us which ideas were impractical or why they wouldn't be fun," Explains Dehlin.