Masters Thesis

Why Dead-Ends?

It is a fairly common belief in game design that having dead-ends in a level is bad. Tim Willits, of id Software says of first-person-shooter (FPS) deathmatch levels, “Build with as few dead ends as possible-they're [sic] best built with none.” [source] In the context of a multiplayer deathmatch game this idea is true, though it applies to single player games as well. Games like Quake 4 (Raven Software 2005) and Halflife 2 (Valve 2004) contain several dead-end areas as part of their single player campaigns. These dead-ends serve to advance the plot while making the game experience seem more realistic by changing the level as the player backtracks. Quotes like these however do not address non-FPS style games, but the ideas and reasoning hold true across several game genres.

The concept of a “sandbox” game is another trend in the game industry. These types of games utilize open ended worlds where the player is free to explore with little guidance from the designer outside of small goals. Because these types of games rely on giving the player the feeling that they can go anywhere, they generally have very few dead-ends. If the “sandbox” games are any indication of the game industry it is that developers see dead-ends as an element which does not contribute to good gameplay. Sandbox games like Grand Theft Auto 3 (Rockstar 2001) and its amazing sales show that a game without any dead-ends can be fun, but evidence that leaving out dead-ends leads to fun gameplay does not prove that dead-ends themselves are harmful to gameplay.

So are dead-ends the bane of all level design? Or are they just another technique that is frequently misused and misunderstood? This project addressed these concerns as well as others by examining player responses to dead-end gameplay and different approaches to making dead-end gameplay fun.