An Isometric Trading RPG game where you explore and survive the harsh, open-world desert. I did the level design for the game, using rational level design combined with 360 open-world from Ubisoft and the mission design of Valve. I designed the world from concept to release, bringing all our features together and designing half of the play space.
UE 4.27 Jira Project Planning 15 Persons Team 8 Months
Development Process
Concept Research
As a starting point, I researched the reference games similar to our creative brief such as Sid Meier's Pirates! (2004), Windward, and Sunless Sky/Seas, (re-)playing them throughout development. I made research documents and summarized them in digestible presentations for my teammates. Examples shown below are excerpts of my research on environmental challenges. Through research, I have found the gameplay of the reference games lacked the most in the traversal element of the game. My focus was to prevent this in Dune Strider by adding interesting and engaging environmental challenges for the player.
I did research on level design for open-world games as I never worked on one before, and found great resources for it, from Iuliu Cosmin-Oniscu's breakdown of open-world level design and outpost designs covering Variety Matrices, Pascal Luban's open-world design purposes and what they're trying to achieve, and molecule design used by Valve on Half-Life called "Experiential Density".
Plenty of findings overlapped with one another. Ubisoft designers reference the 40 Second Rule and Nintendo has a Discovery grid which serves the same purpose, the player should encounter challenges, resources, or otherwise, be engaged on a higher level throughout gameplay.
The 360 Approach Design of Far Cry and Assassin's Creed (and other Ubisoft open-world games) offers multiple options for the player to engage with every mission/outpost, similarly covered by Pascal Luban's breakdown of the main difference between open-world levels as opposed to preset maps or areas, and that choice is the core of open-world games.
Following the research I've done, leads placed us in strike teams to converge on ideas that the team decided were most within the scope of the project and in line with the creative brief we had. As part of my Environment Challenges research, I joined a strike team that made a one-pager to serve as inspiration and direction for how we saw environmental challenges and other level design ingredients.
As I finished my open-world level design research, I condensed it into a proposal as part of a sprint review for the team, in preparation for the pre-production stage of development:
Pre-Production
I started with establishing high-level pipelines and documentation for the rest of the development, condensing research conclusions as the reference for concept work.
During this stage, we had issues within the team with communication, and our small team of 4 artists decided to part ways with us, leading us to use art asset packs from the marketplace. Once leads chose an asset pack that works with our custom shader and the visual direction of our game, I created a showcase art scene to understand how easy it is to use it, how long it would take to create a point of interest and what we need more to make it better.
My main discovery is that there were too few buildings in the asset pack for the number of factions and variety our concept demanded, and I discovered issues with VFX and our shader not working properly that programmers quickly started to tackle.
Following this, I was working on the proof of concept and first playable demo that teachers would give feedback on.
As the Pre-Production stage was nearing its end, we had no one managing documentation for quest implementation so I made the first draft using research and proposals made before me by our lead game and narrative designer, and established the structure for others to use.
Production Stage
We divided our work, and on my initiative we used the World Composition feature of UE,
to improve workflow by avoiding bottle-necking with source control checking out of levels.
I made the (albeit crude, due to the short time we had) map where we would work on separate grid parts and each of them was either a transition area, landmark, settlement, or harvesting location.
This helped us have an overview of work done and remaining, as well as easier to iterate upon it following QA and playtesting conducted. I designed, iterated, and set-dressed half of the playspace.
From early in-house testing and QA all the way through the production and release stage, iterations were introduced to the play space I was the level designer of. Halfway through development, we noticed that players were still struggling to orient themselves, so I suggested we request from the tech art outsource team for a road tool that would help us guide them more, as well as include smaller but recognizable landmarks in every grid.
Conclusion
As a team, we encountered great struggles and misunderstandings of what our vision was, but we managed to deliver a product that was above expectations. I greatly enjoyed working with the tech art outsource team, and I am satisfied with my work on the game and have learned substantially from this project.
As a level designer, my biggest struggle was guiding the player through an isometric game without a map. Later in development, I requested and introduced the road tool and it helped with guidance, and system designers included UI markers as well, helping with guidance. We also tried to include a bit more complex map feature, but it was scrapped due to lack of time. Learning how to set-dress, visual composition and lighting was great for exploration and enhancing my design capabilities and it benefited me in the long run.
コメント