From the tf2 blog:
We thought you might be interested in seeing some of the "behind the scenes" work that goes on in Team Fortress 2. The process usually starts with a team discussion: This helps us determine the goals of the map we're going to build. In this case, we wanted a game mode that had an intense, centralized experience with quick player turnaround. We liked some of the facets of Arena mode, but it was clear that many players don't like to wait for the next round to play.
The sooner we could prototype the new mode, the sooner we could start gathering data on it. To this end, we converted some of the existing arena maps and started experimenting. Some were more successful than others. We quickly found out that fun arena maps weren't necessarily fun KOTH maps. Playtesting the new mode on existing maps helped us explore various ideas that we used to generate rough sketches of what would become Viaduct. In order to get some real data on its success or failure as fast as possible, we used Hammer to quickly turn the sketches into a very rough version of the map. Then we started playing.
Here's a timelapse video showing the evolution of Viaduct:
-Video-
Here's another that shows the map coming together as a whole:
-Video-
We playtest daily. Almost every version you can see in the timelapse video was at one point tested with a mix of experienced and new players. That way, decisions are driven by the observations and feedback generated by the test. As an example, at one point several sessions in, it felt as though the player wasn't being given enough of a "cool-down" period after being in the action at the control point. The pacing was actually too intense. In addition, teams that captured the zone felt they had no space in which to keep control, due to the continuous stream of enemies pouring into the space. This devalued some classes significantly, like the Engineer. We were able to address both of these issues through the creation of an area in front of the respawn rooms. This allowed us to push the spawn points further away from the control point.
Once the map started progressively improving in terms of playability, we started to feel more confident about the new game mode. Now the art pass could begin. We were still playtesting the map throughout this phase, though the layout changes tend to be on a smaller scale (such as moving health and ammo, adjusting jump distances between objects for classes, and addressing sniper sight line issues). The snow, as you can see from the timelapse, was a late addition—the idea was put forward to extend the alpine theme and provide a new, visually interesting space to fight in.
Link - http://www.teamfortress.com/post.php?id=2754
We thought you might be interested in seeing some of the "behind the scenes" work that goes on in Team Fortress 2. The process usually starts with a team discussion: This helps us determine the goals of the map we're going to build. In this case, we wanted a game mode that had an intense, centralized experience with quick player turnaround. We liked some of the facets of Arena mode, but it was clear that many players don't like to wait for the next round to play.
The sooner we could prototype the new mode, the sooner we could start gathering data on it. To this end, we converted some of the existing arena maps and started experimenting. Some were more successful than others. We quickly found out that fun arena maps weren't necessarily fun KOTH maps. Playtesting the new mode on existing maps helped us explore various ideas that we used to generate rough sketches of what would become Viaduct. In order to get some real data on its success or failure as fast as possible, we used Hammer to quickly turn the sketches into a very rough version of the map. Then we started playing.
Here's a timelapse video showing the evolution of Viaduct:
-Video-
Here's another that shows the map coming together as a whole:
-Video-
We playtest daily. Almost every version you can see in the timelapse video was at one point tested with a mix of experienced and new players. That way, decisions are driven by the observations and feedback generated by the test. As an example, at one point several sessions in, it felt as though the player wasn't being given enough of a "cool-down" period after being in the action at the control point. The pacing was actually too intense. In addition, teams that captured the zone felt they had no space in which to keep control, due to the continuous stream of enemies pouring into the space. This devalued some classes significantly, like the Engineer. We were able to address both of these issues through the creation of an area in front of the respawn rooms. This allowed us to push the spawn points further away from the control point.
Once the map started progressively improving in terms of playability, we started to feel more confident about the new game mode. Now the art pass could begin. We were still playtesting the map throughout this phase, though the layout changes tend to be on a smaller scale (such as moving health and ammo, adjusting jump distances between objects for classes, and addressing sniper sight line issues). The snow, as you can see from the timelapse, was a late addition—the idea was put forward to extend the alpine theme and provide a new, visually interesting space to fight in.
Link - http://www.teamfortress.com/post.php?id=2754