Postmortem

What went right:

I built and tested a prototype for the water rising mechanic during the level concept stage. By doing this I knew how the level’s unique mechanic would work, and had an idea how it would affect gameplay. I was able to write the level design document with this knowledge in mind.
Since I had started with an early prototype, I was able to build several versions of the map during the whitebox stage. This meant that I was able to work out a few of the issues with level flow and refine the water rising mechanic.
While building some of the initial Matinee cinematic effects for the riverbanks, I realized that the apparent motion suited a ship sailing down a river more so than a building being washed away by a flood. Since this happened late in the whitebox phase of production, I was able to change the level’s aesthetic and remain on schedule.

What went wrong:

As water engulfs the sinking ship, lighter objects should begin to float. This created an issue because those objects affected gameplay, and Unreal Tournament 3 did not include dynamic path nodes which would allow me to adjust the AI’s navigation around those objects. I ended up replacing the more interesting floating objects with containers that were restrained by large cargo doors.
Playtesters were having difficulty escaping the rising water from several areas of the ship, including the lower level cargo hold. Escape routes had to be added to every area where players might become trapped, complicating the flow for the level.
I decided to change the aesthetic midway through the project. While I was able to handle the change well, it still was a significant redesign that I should have been able to better predict.

What I learned: