Designing Large-Scale Mobile Game Cities: Production, Optimization, & Worldbuilding Expertise

Can you tell us about your background and how you got into world design?

I’ve been working in the game industry for over 11 years, primarily focusing on environment and world design. Over time, I intentionally shifted toward large-scale systems, where design decisions affect not only visuals but also player behavior, performance, and scalability.

My expertise developed at the intersection of design and technical constraints. I approach worldbuilding as a system rather than a collection of assets – designing spaces that guide player navigation and shape user experience, especially in multiplayer environments where player behavior is unpredictable.

What was the core challenge of designing a 20 km² multiplayer city for mobile?

The core challenge was building a system capable of supporting thousands of players moving simultaneously within an open world, while staying within strict mobile constraints in terms of memory and performance.

Unlike single-player games, where player movement can be guided, I had to design a world that naturally organizes player flow.

To achieve this, I structured the environment around clearly defined points of interest. At an early stage, I defined over 200 POIs, including businesses, workplaces, hubs, and key landmarks, which allowed me to establish the functional logic of the city from the beginning and avoid chaotic content placement later in production.

How did you approach structuring such a large environment?

I divided the city into distinct districts – central, industrial, residential, and historical – each with its own functional and visual identity.

I personally defined:

• spatial hierarchy
• content distribution rules
• density across districts
  
  

This ensured readability of the environment and predictable navigation, even with high player density.

At an early stage, I created a full blockout of the city to validate scale, proportions, and player perception before production. This allowed me to avoid costly rework later.

Additionally, I deliberately oversized streets and open spaces. This improved player movement and reduced friction, especially in high-density multiplayer scenarios.

Streaming is a major challenge in large worlds. How did you handle it?

Streaming was one of the core systems I worked with and optimized as part of the project.

Since loading the entire world on mobile devices was not feasible, I structured the environment using Unreal Engine’s built-in streaming system, organizing the world into a grid of streaming chunks. Each chunk contained geometry, props, and textures, and was loaded dynamically based on player position.

Within this framework, I defined and implemented rules for:

• loading only nearby chunks
• preloading neighboring areas
• using multiple LOD levels
  
  

This allowed me to minimize visible pop-in and maintain the perception of a seamless world.

In addition, I established strict memory budgets per chunk, including limits on object count, materials, and textures, which ensured stable performance even in dense urban areas.

What optimization strategies did you implement?

I worked directly on several key optimization areas:

LOD and Geometry
I implemented multi-level LOD systems and carefully tuned transition rules to minimize visual artifacts. For distant objects, I introduced impostors and billboards, preserving visual density while reducing rendering cost.

Draw Call Optimization
I implemented batching and texture atlas strategies, significantly reducing draw calls and stabilizing performance. I also used instancing for repeating assets such as buildings and infrastructure.

Asset Standards
I contributed to defining production-wide guidelines, including polygon budgets, material limits, and shader reuse. This ensured consistent performance across the entire project.

You also contributed to pipeline development. Can you elaborate?

Yes, pipeline development was a key part of my contribution.

I introduced a production system based on photo-derived textures, which allowed us to scale efficiently while maintaining visual realism.

Within this system, I:

• contributed to the creation of over 4,000 assets
• helped build a reference library of more than 25,000 real-world photos

This significantly accelerated production speed, reduced asset preparation time, and improved visual consistency across the project.

How did you structure the environment production?

After validating the blockout, I led the transition into full production.

I structured the workflow into clear stages:

1. Terrain
2. Architecture
3. Environment detailing

This allowed me to maintain control over scale, accelerate production, and minimize late-stage errors.

All production was built on top of the blockout, ensuring consistency across the entire world.

How did you approach materials and texturing?

I designed a scalable material system optimized for performance.

We used a standard setup:

• Base Color
• Normal
• ORM

I maximized reuse through trim sheets, shared materials, and modular textures. This significantly accelerated production while maintaining visual consistency.

How did you address mobile platform constraints?

I defined and implemented technical standards specifically for mobile platforms.

These included:

• texture resolution limits (typically 1024×1024)
• ASTC compression
• aggressive texture reuse
• atlas-based material systems

These decisions allowed me to balance visual quality with performance, ensuring stable FPS even in complex scenes.

What do you think is the key to building large-scale multiplayer environments?

The key is a systems-driven approach.

I treat world design, optimization, and production as a unified system, where every decision affects scalability.

For large multiplayer environments, it’s critical to have:

• structured world design
• predictable systems
• strict technical discipline

This is what enables scalable and stable environments.

What are your future plans?

I continue to focus on large-scale environments and system-driven world design, building on my experience in creating scalable multiplayer worlds.

I’m particularly interested in advancing:

• high-density multiplayer spaces
• scalable production pipelines
• deeper integration between design and technical systems

At the same time, I’m actively exploring how AI can be integrated into worldbuilding workflows. My goal is to develop new methods that combine procedural systems, automation, and AI-assisted tools to significantly accelerate the creation of large-scale virtual environments without compromising quality or design intent.

I see strong potential in redefining how complex game worlds are built – especially as technical demands continue to grow and production cycles become more demanding.

Recommended Reading

• The Image of the City by Kevin Lynch
• The Design of Everyday Things by Don Norman