In the second year of my bachelor's study, I did a level design project to create a Blitz game mode level for the latest Unreal Tournament game. The project taught me the proper processes and pipelines for level design: From sketching and concept proposal that supports the core gameplay of the (new) game mode to QA, playtesting, and iteration processes.
UE 4.27 Jira Project Management Solo Project 8 Weeks
Level Design Process
Research
Played the franchise and established a research document about the IP with clear research questions. Following that, I researched the Blitz game mode and analyzed an existing level (exclusively made for the mode).
Through research, I established level design metrics, such as distances/time between points of interest (such as chokepoints, checkpoints, and weapon and power pick-ups), as well as the groundwork for 3Cs (speed, size, and so forth). I also found level design rules and guidelines for the Blitz game mode from Epic Games' level designers.
Researched and interviewed target audiences, their gaming profiles, and what they seek from this type of game (and the game mode).
Concepting
As an exercise, we took reference pictures that we explored (or just used directly) by painting over them for inspiration of our level.
Following what we learned through our research, I used SketchUp for the first time to quickly build and get feedback on my concept. After numerous iterations, I started working on a concept proposal.
After a couple of concept models and feedback sessions with fellow students and teachers, this is the concept model of my level. It follows almost all core rules and guidelines found through research, with few exceptions for the sake of iteration and curiosity regarding the gameplay that I wanted to explore through playtesting to see how well it would work.
Defining the theme of the level helped a great deal with how I approached the visuals of the level while establishing level design pillars, alongside rules and guidelines found in Epic Games documents, kept my design on track, creating a design-sound level.
Concept Breakdown & Grayboxing
After presenting my level concept, I received a great deal of feedback and immediately iterated upon it. I created a more concise level breakdown. Feedback from veteran players was helpful as I realized my level is far too open, with too many lines of sight that both sides could exploit.
While implementing the level in-engine, I realized I thought little of lighting and of the edge-of-world (mainly the roof, as was pointed out by my teacher when I asked for feedback). I decided to implement rudimentary lighting, and started working on the edge-of-world, and how it will be utilized gameplay-wise. For example, I envisioned in the Sacrifice Altar room, the broken pillars from the ceiling were used as cover for a higher floor.
Iterating
Once I started iterating, I realized while testing it with fellow students that despite seeing the symmetry of the level as a challenge imposed by the teacher, it was not working as the entire game mode was designed around asymmetry in mind, to offer switching attack routes for attackers to outmanoeuvre the better-positioned defenders. To the right, you can see all the major iterations that I've documented over the 8 weeks.
With asymmetry in mind, version 0.16 added two experimental paths that I tested with fellow students and veteran players to figure out why they prefer each one and what they use the most inside of those areas. I also paid close attention to what they found difficult to play around with, and where they were blaming the level for consistent deaths and overly hard-to-defend/attack areas. Iterations went from huge changes to more minute ones as the project went on, as I managed to find a decent balance for the level the more I worked on it. The final notes speak of what I would change if I had more time, but the project ended there (version 0.21) and I was satisfied with my work.
Conclusion
I was thrilled with my work, and my progress as a level designer. I learned a great deal about upkeeping a clean workflow, having a clean world outliner, and interpreting feedback from players and properly posing questions for it, developing my QA skills further.
I also used many tools for the first time, such as SketchUp and the Brush tools in UE. If I could start over, I would've scrapped the level and made it much smaller, as I realize that the scale, although wanting to be grandiose and very similar to old levels from UT99, was far too large and open, and most of my iterations went on fixing that instead of starting from scratch at an earlier phase. I would also incorporate more visuals to distinguish the areas, as I kept it strictly gameplay-focused and didn't spend time on shaping the world beyond that, and I think that would've helped the players (besides level geometry) to be able to quickly understand their position and options. Nevertheless, iterating has brought me to a much more satisfying level that seemed balanced. Players were giving positive feedback and a win-loss ratio and other testing proved that the level is balanced and well designed.
It was a difficult challenge as this was a game mode that was new to the UT, and it felt like I was working for the company and figuring out how this game mode can be fun and engaging.
Comments