Microsoft's Interactivity Lab

There is an interesting article in last month's issue of Wired that described the various techniques and facilities employed by Bungie to improve the Xbox 360 game Halo 3 as compared to its predecessor. The article reminded me of one reason that game development is so fascinating to me. So many different areas of expertise are tapped to develop a game - computer science, visual arts, music, writing, and, increasingly, other fields such as psychology and cognitive science. (As an aside, I think at least an introductory background in HCI would be helpful to all members of a team involved in the cognitive aspects of game development.)

Unfortunately, level designers are not players (at least, not in this context). They are too close, intellectually, to be certain that the intended progression through a level follows from its design; players may be oblivious to the "proper" path that seems glaring to those who designed it. Confusing or obfuscated design frustrates players, weakens the game's narrative, and in the worst case can cause players to quit altogether. To combat this problem, the team at Bungie wrote a tool that took note of a player's position at one-minute intervals and overlaid corresponding points onto a map of the level. The resulting aggregation of several players' runs showed that the level in question was confusing: Point patterns showed that players were routinely traversing the same paths, traveling in circles! Armed with this insight, the level designers altered the map to better direct players by changing the layout of "accessible areas" and adding visual cues, such as the arrangement of foliage. After the redesign, the team ran the test again. The new pattern of points showed a more regular stream throughout the level, demonstrating that players were now consistently able to find their way within a reasonable amount of time.

This article shows why observing how a user interacts with a system is such an important step in making that system satisfying to use. Bungie's observations tested many of the assumptions about player behavior made by Halo's developers, showing that several of them were indeed false and would have detracted from the quality of the game if left unchallenged.

It's probably safe to say that a good portion of the Bungie team consists of gamers. I find it interesting that even gamers, when acting as designers, don't know for certain what will work for other gamers.