Cultist Castle
Unreal, First Person Shooter
Cultist Castle
Unreal, First Person Shooter
Cultist Castle
Unreal, First Person Shooter
Drosera - "Large" Team Development
Drosera Project
Drosera is an isometric, procedurally generated, dungeon crawler set on a sci-fi/cyberpunk alien planet. A team of up to two players choose from two character classes to begin an expedition onto the alien landscape to capture a rare plant called a "hyper-seed" to fuel their space ship. The native alien species are plant-like sentient creatures that defend their natural environment from the encroaching threat of humans exploiting the natural resources of this world. The game plays out over a series of 5 "levels" of increasing difficulty, where each level generates a procedural dungeon environment out of various room types, with successive levels having a more difficult arrangement of rooms.
​
Drosera Team
Drosera is a follow-up to a previous UT Dallas Game Lab project I worked on, called Sciophobia. Due to issues and conflicts that developed due to the corona virus quarantine in spring 2020, that project was prematurely cancelled. I applied to work on another Game Lab project for the fall 2020 slate. This project was my opportunity to finish what I had started in the spring, with many of the team being the same people, I see Drosera as a mulligan on the Sciophobia development experience.
​
The game was pitched by the same creative director as Sciophobia, Ellie Goyer, and I was excited to work with her again on a much larger team in the fall of 2020. My role as Level Designer on Sciophobia carried over to the new project, and I was promoted to "Lead Level Designer" with two additional designers working under me, James Dollar and Sydney Calvert. Together our team collaborated closely with the Game Design team, headed by Aiden Montgomery (also from Sciophobia). Both design teams we were directed by our project manager, Charlotte Royal. The roles we had working on Drosera also included a programming department, an art department with teams for environment art, character art, animation, and more, as well as a VFX team, a UI/UX designer, a sound department, and a few other individual roles that altogether formed a development company of over 40 team members.
​
The Game Lab was lead by two faculty members from the UT Dallas ATEC department, professors Adam Chandler and Timothy Lewis.
​
Drosera Development
This project was the largest scale and scoped project I have ever worked on, and I was thankful to have an important role as Lead Level Designer to help input some of the creative direction for this project. I set forth the expectations and procedures for my team to follow as well as giving feedback and guidance on a close level for both of my designers. Our work was closely directed by our department manager and we had helpful guidance from the faculty at UT Dallas.
Final Level Generation Tool
Level Generation Tool
I worked with a small team to design the procedural generation tool for Drosera. Along with a designer from the Game Design team, I worked with members in the Programming Department to draft out the generation algorithm.
​
The basic loop that the tool follows is as such:
The level is assigned a random "biome" (gameplay implications here), and begins with a "dropship room." The dropship is standard at the start of every leve, and equally a "hyperseed" is always at the end. After the dropship, another room is selected at random and stitched together at the doorways. Each room is defined as a particular set of geometry, and has multiple "layouts" of enemies and pickups. Each layout has a defined "difficulty," and as each room & layout is generated the difficulties are added together until the level's "difficulty cap" is met (or supassed) and generation stops. Then the hyperseed room is generated and the level is complete.
​
Narrative Consultation
Together with the Narrative Designer (@Brittney Kopel), the Level Design team constructed a spreadsheet of narratively appropriate locations to theme the rooms around, as well as enemy encounters that we would like to see mechanically represented. Whenever we found a matching pair, we would iterate on the concept and post the final room pitch. Each designer was then responsible for 6 of the rooms, and put together an asset list of enemies, interactables, and important art assets that we wanted to include.
This is also where tentative difficulty was decided for the rooms, with flexibility in mind for game balance and variations in expected encounter designs (layouts).
Narrative Consult Document, per Room/Level
Papermaps
Each designer was responsible for 6 rooms, for me that meant the initial Dropship room, the end-goal Hyperseed room, and 4 additional rooms. The rooms were assigned to us with various difficulty benchmark ratings, which are integrated into those room's names (i.e. 04b_WaterHole is the Seocond (b) difficulty 4 (04) room).
​
My initial papermaps have all been adjusted during whiteboxing and playtesting, but the initial concepts for each room holds true. I gave my designers a style guide .psd file so that all of our drawings would have at least the same color scheme and legends to ensure we could all quickly understand each other's concepts and give feedback.
My "Rooms"
Screengrab of Narrative Consultation Doc
All of my Papermaps for Drosera
All of my Whiteboxes for Drosera
Whiteboxing
The basic whiteboxes were submitted for each room before our Alpha deadline, 2 weeks after papermapping was complete. Each room required certain technical requirements, such as using Unity's NavMesh system, having special inspector hook-ups for assets like "entrance marker" and layout groups.
In-Editor View of Room 04b_WaterHole Whitebox
Art Integration
We (level designers) had made ourselves a small problem when creating our assets in Unity, as we did lot leave enough wiggle room for the art teams to implement within our nested prefabs structure. I (as the lead) had to reconfigure the scene hierarchies of each room, as well as reorganize the folder structure for our assets, so that we left ample room for the various art people to work simultaneously without running into any merge errors down the road. The changes were relatively minor, but boiled down to creating a new parent object to hold our collisions, and making the art on the same level as the collision, instead of being child objects.
​
Once the art teams had their go, with some guidance from the designers to spell out some of the gameplay intent in our levels, the results were rather spectacular. All of the rooms had two art passes, reflective of the two biome types available in the game.
Each of my rooms, with either their Jungle or Desert environment art passes
Final Build
After guiding the artists through their art passes, and continuing the ever fruitful pursuit of bug-fixing, we eventually reached content-lock on our project. I believe I had minimal to no gameplay related bugs with my rooms, and perhaps just some art-collision mismatches, and I guided the rest of my level design team to fix their bugs (if any). By the end of the project, I'm pretty happy with my rooms and my overall contribution to the team!
Final Gameplay Footage, Courtesy of UT Dallas