low poly

LOW POLY VS HIGH POLY: WHICH IS BETTER FOR GAMES

The polygon count debate in game development sounds like it should have a simple answer. More polygons equals more detail equals better graphics equals better game. In practice, every part of that chain is wrong.

More polygons does equal more detail. But more detail doesn't equal better graphics. Better graphics doesn't equal better game. The relationship between polygon count and game quality is complicated, contextual, and frequently misunderstood.

What high poly does well

High-poly models capture detail that low-poly models can't. Wrinkles on skin. Fabric folds. The curve of a doorknob. When the goal is photorealism, high polygon counts are necessary because the human eye can detect even small geometric imperfections. A face with 10,000 polygons looks like a face. A face with 100 polygons looks like an interesting geometric abstraction.

For games that trade on visual spectacle, high poly is essential. Nobody wants to explore a photorealistic open world where the trees are hexagons. The Last of Us needs high-fidelity faces to sell its emotional performances. Flight simulators need detailed cockpits to maintain immersion.

What low poly does well

Low poly communicates. A low-poly character with 300 polygons can be instantly recognizable if the shape language is strong. Superhot's red crystalline enemies are more visually distinctive than most high-poly character designs because the simplicity forces clarity. You know what they are immediately. No visual noise competing for your attention.

Low poly also runs on everything. A low-poly game that targets 60fps on mid-range hardware will reach a larger audience than a photorealistic game that stutters on anything below a 4090. Performance is a design choice that affects accessibility.

Production time is the practical argument. A single high-poly character for a AAA game takes weeks of artist time across multiple specialists: concept, modeling, texturing, rigging, animation. A low-poly character can be created by one person in a day. For small teams, this isn't a nice-to-have. It's the difference between shipping and not shipping.

The uncanny valley problem

High-poly games that aim for realism but don't quite achieve it fall into the uncanny valley, where characters look almost human but something is off. The eyes are too glossy. The lip sync is slightly wrong. The skin doesn't scatter light correctly. The result is a visual that's worse than either full realism or deliberate stylization.

Low-poly art sidesteps the uncanny valley entirely. Nobody expects a 300-polygon character to look human. The abstraction creates a different relationship between player and character. You project humanity onto the simple geometry rather than evaluating it against a realistic standard.

When to use each

High poly is the right choice when realism serves the game's goals. Horror games benefit from realistic environments because familiar spaces that look "real" feel more threatening when something goes wrong. Narrative-driven games benefit from realistic faces because emotional performances require facial detail.

Low poly is the right choice when clarity serves the game's goals. Strategy games benefit from readable unit silhouettes. Arcade games benefit from instant visual parsing. Any game where the player needs to process a lot of visual information quickly benefits from the reduction of visual noise that low poly provides.

The middle ground

Most modern games don't sit at either extreme. They use stylized art that falls somewhere between low and high poly. Fortnite's characters are detailed but not realistic. Hades uses painterly 2D art for portraits but simplified 3D for gameplay. Valheim shipped with low-res textures on medium-poly models and became one of the biggest survival games ever, because the art direction made those constraints feel intentional rather than cheap. The choice isn't binary. It's a spectrum, and each point on the spectrum has different tradeoffs. A game with medium poly counts and strong art direction will almost always age better than a game that chased photorealism with the hardware of its era. Look at Wind Waker versus any realistic GameCube title. The stylized middle ground wins the longevity argument every time.

Why I build in low poly

I build low-poly games because the constraints produce better design decisions. When every polygon matters, you think harder about shape. When textures are minimal, you think harder about color. When detail is limited, you think harder about what to emphasize.

The constraint forces intentionality. A high-poly modeler can add detail everywhere and let the player's eye find what matters. A low-poly modeler has to decide what matters in advance and allocate polygons accordingly. That decision-making process produces art that communicates more effectively per polygon than any amount of brute-force detail.

Low poly isn't low quality. It's a different set of priorities. And for the kinds of games I make, those priorities produce better results.

← Back to the Sketchbook