The Short Version

Evacuation was released on the XBox Live! Indie Games channel in September, 2010. It's a single player action/strategy game where you can use a wide variety of tools and vehicles to hold off mutant hordes while you evacuate as many people as you can from the city.

We were a small, four person team that did this over the summer for fun. We also wanted to learn more about the process of building a console game. Since we knew we'd be completely overwhelmed when school started again, we had a strict time limit of two months.

I primarily served as the team's 3D artist and texture artist. Though as is often the case with small teams, we all ended up doing a little of everything. My other major contributions were some rendering optimizations, minor special effects shaders, and a handful of scripted tools and animation systems that unfortunately did not make it into the final game.

Unfortunately the end result is a mixed bag. The gameplay is very fun, but there are many fewer levels than we wanted, and the visual quality is lower than we hoped. Visually, I'd say it is on par with a first stage blocking pass, because due to time constraints we really only had the opportunity for a quick draft.

Our vision for this game far exceeded our time and our abilities, but we wouldn't have it any other way. If we aimed for "easy" or "comfortable", it would have been a total waste of a summer.

The Long Version: A Post-mortem of Evacuation's Visuals

So, I'll just get it out of the way. Evacuation's graphics are not amazing. As the lead artist, that's on me. Most of that is due to my inexperience as a texture artist. We needed very simple building models, which means that a lot of the visual quality is then left up to the textures. Good textures can make simple models look fantastic. Or in our case, it can make perfectly serviceable models look "kinda bleh", scientifically speaking.

A core concept that we wanted to maintain was the feeling that the player was in a large city environment that was full of people that needed to be saved. This meant that, ideally, we'd have a ton (metric or standard) of buildings that the player runs around while saving hundreds of civilians from a never-ending stream of mutants. Sounds great, but it's quite a challenge.

As the team's only full-time 3D modeler it was up to me to model and arrange all of the buildings for all of the levels. Initially we started with Google SketchUp since it's easy to make simple buildings very quickly. I could churn out hundreds of unique buildings, but the problem was that I'd need to texture them all separately.

Unfortunately, I'm just not that fast at unwrapping and texturing. So instead, I decided to model and texture a handful of building types that could be copied multiple times and spread around each level. I could make a few color swatches for each building type, and then we could rotate and scale them appropriately all across the maps to vary up the environment as much as possible. In retrospect, there should have been more buildings and more textures for each one, not just color swaps. That would have helped a lot.

We also changed visual styles mid way through due to frame rate issues. We had intended for there to be hundreds of fully animated 3D characters running around the scene. This was naive, but it would have been so awesome! We kept scaling things back and back, but we absolutely needed hundreds of people on screen for this gameplay to work. We tried it with dozens, but it really just didn't work the same way. Maybe it's clever, maybe it's lazy, but what we did was start styling the game as if you were a remote agent who was doing this all through a satellite interface. All the people would be identified by special 2D markers that would signify their current status. Green was good, gray was dead, and blood splatters meant horrible, horrible things had just happened. In the end we wound up with a mixed style that didn't have the cohesion it needed for this concept to really take off.

Still on the TODO List

There were a handful of things I wish I would have been able to get to, but alas, they were too far down on my priority list. For example, I intended to make unique buildings for each level. I also wanted to implement a more advanced lighting system that would simulate (either through textures or through post processing) indirect illumination, and a soft shadow mapping shader to give the game a better sense of depth. I would have been overjoyed if we could have gotten my rigged 3D character models to work at the capacity we needed. But unfortunately, these are all things that are left undone. This leaves the final visual quality looking like a rough draft, first pass, conceptual blocking sort of stage. I also wish that I'd gotten a chance to redo the User Interface, which was drawn and coded by our lead programmer. It was one of those "build it while we go because we need it now" sorts of things that just ended up staying there. We just didn't have the time to do it right.

Overall, it's a bit embarrassing, visually, because I know that with just one more month I could have made it look much better. But when it really comes down to it, I'm proud of this game. Evacuation is a lot of fun to play and it was a great experience to make it.

Back