Why Most 3D Game Advice is Wrong

Why Most 3D Game Advice is Wrong

The internet is flooded with advice for game developers, particularly regarding 3D development. From "always use normal maps" to "never use vertex colors," these tips are often presented as absolute truths. However, the reality of shipping successful titles proves that rigid adherence to outdated or overly simplistic rules can actually hinder a project's success.

The Myth of Normal Maps

A common piece of advice is to always use normal maps for terrain and surfaces because they provide high-fidelity detail without adding geometry. While this is true in theory, it fails dramatically when applied universally.

• Skyboxes: Using a normal map on the skybox creates an unnatural, wavy appearance that breaks immersion immediately.
• Foliage: While foliage often benefits from displacement, many trees in shipped games rely purely on vertex colors to define leaves versus branches. Displacement would overdraw these surfaces and kill performance rather than help it.
• Volumetric Light: Modern techniques like volumetric god rays are increasingly implemented by painting a low-res normal map onto the skybox plane in post-processing, proving that normals can be effective for atmospheric effects when used correctly.
• Hair and Fur: Hair shaders almost never use displacement or normal maps; they rely on vertex colors to define individual strands.

The lesson is clear: context dictates the tool. Blindly applying a technique because "it's good" without considering the specific surface type leads to visual artifacts rather than optimization.

The Danger of Fixed Asset Sizes

Another piece of advice is to never resize game assets in code, suggesting that designers should manually pack everything into perfect powers-of-two dimensions. This is dangerously inflexible for modern development cycles.

• Tile Variations: In games like Cities: Skylines, a single city tile needs to support different sizes of buildings, trees, and decorative items. Designers cannot pack these into fixed slots because the variety required is too great.
• Multiplayer Syncing: In multiplayer titles such as Ants io, every asset must be packed into identical sized tiles so that client-side rendering matches server-side logic. This strict requirement often forces developers to use larger, inefficient tile sizes even when smaller ones would suffice.
• Memory Efficiency: Assets are often stored in compressed formats where memory usage is calculated as width × height. Picking an odd number like 513×513 wastes exactly half a kilobyte compared to choosing 512×512, yet many tools fail to warn users of this inefficiency.
• Lod System Compatibility: While Level-of-Detail (LOD) systems require consistent sizes, they do not mandate that the base LOD match perfectly; smaller assets can be scaled up in code without issue.

Rigid sizing constraints often lead to poor performance and bloated memory footprints. Flexibility allows designers to create richer content.

The Fallacy of Universal Shaders

Developers are often told to avoid writing custom shaders, preferring only standard engine materials. This ignores the massive variety of lighting setups found in shipped games.

• Lighting Compatibility: Standard engines like Unity have default "Unlit" shaders that do not interact with studio lighting (like HDRP), causing black objects in lit scenes.
• Ocean Shaders: The standard ocean shader is notoriously bad and difficult to tune without custom code, making it nearly impossible to get a realistic water surface out of the box.
• Character Passes: In games like Dota 2, certain passes (such as hair) must be completely opaque or transparent; standard shaders often struggle with these specific requirements without custom tweaking.
• Baked Lighting Issues: Custom lighting is usually baked into static meshes rather than calculated dynamically, meaning a simple diffuse shader will not look good in the final product.

The industry has moved away from rigid material systems because they cannot handle the nuanced lighting and artistic requirements of modern 3D games.

A Self-Teaching Framework for Developers

If you find yourself doubting the advice you are reading, apply this self-teaching framework. It forces you to validate information against real-world shipped titles rather than accepting dogma blindly.

1. Identify: What specific claim is being made? (e.g., "Normal maps are always best for terrain.")
2. Hunt: Search your memory or research for shipped games that contradict this claim. Look for titles known for high fidelity.
3. Verify: Does the counter-example actually exist in the wild? For instance, does a popular game use vertex colors on leaves?
4. Conclude: Formulate a nuanced takeaway that explains why the original advice was wrong or limited.

Key Takeaways from Shipped Games

To illustrate this framework, consider how specific titles handle their graphics pipelines. The following table outlines key technical decisions in three popular games:

Game Title	Key Asset Handling Strategy	Why This Matters
Cities: Skylines	Dynamic asset packing	A single tile supports varying sizes of buildings, trees, and decorations.
Ants io	Strict uniform tiling	Assets are packed into identical sized tiles to ensure multiplayer synchronization between clients and servers.
Dota 2	Cross-platform rendering	The same asset is rendered in both high-end and low-end contexts, requiring careful shader management.
Skyrim	Dual-resolution approach	High-resolution assets are used on PC while lower-res versions run on consoles or mobile devices.

The second table demonstrates how these games specifically handle normal maps, showing that usage is highly context-dependent:

Game Title	Terrain/Building Treatment	Vegetation/Light Treatment	Hair/Fur Treatment
Cities: Skylines	Uses displacement	Mixed approach	Volumetric normals on skybox
Ants io	Vertex colors for leaves	No displacement	Pure vertex colors
Skyrim	Mixed approach	Displacement for grass	Vertex colors
Dota 2	Volumetric normals on skybox	No displacement	Pure vertex colors

👉 Played enough? Try Why Two-Player Game Advice Often Fails

👉 If you enjoy Why Most 3D Game Advice is Wrong, you will also enjoy Hospital Doctor Emergency — try it here

By analyzing these concrete examples, you can see that "best practices" are not universal constants. They change based on the genre (e.g., the endless highway runner mechanics of games like Runner Tomato: Hyper Casual demand different performance considerations than a strategy game). Always question the source and verify with evidence from shipped titles.

👉 If you enjoy Why Most 3D Game Advice is Wrong, you will also enjoy The Browser Arena Paradox — try it here

👉 Level up your game experience

👉 Played enough? Try State of Casual Browser Games in 2026

👉 Played enough? Try The Browser Frontier: Adventure Games Evolution

👉 Level up your game experience

👉 Played enough? Try Browser Action Games: Design, Business, and Technology

👉 Level up your game experience

👉 Played enough? Try Beyond the Binary: The Flawed Action vs Battle Taxonomy

👉 Level up your game experience

👉 If you enjoy Why Most 3D Game Advice is Wrong, you will also enjoy The Car vs. Driving Binary: A Taxonomy Failure — try it here

👉 Level up your game experience

👉 If you enjoy Why Most 3D Game Advice is Wrong, you will also enjoy Abyssal Echoes vs Special Forces X: Design Constraints in Browser Action Games — try it here