1 Overview
Interviews for game art roles tend to drift toward the same comfortable topics. Candidates want to talk about which renderer they prefer, which sculpting brush pack they rely on, and how many millions of polygons their latest hero character carried before decimation. Those conversations are pleasant, but they rarely reveal whether someone can ship work inside a team. That is why I steer interviews toward retopology instead. Retopology is the moment where an artist stops making something look good in isolation and starts making something that can actually run, animate, and be edited by other people.
The question is not a trick and it is not a gatekeeping ritual. It is a practical filter. A portfolio can hide a great deal behind flattering lighting and a single beauty render, but topology exposes habits. Clean edge flow around joints, sensible density in flat areas, and deliberate decisions about where loops need to sit tell me how the artist thinks about the downstream consequences of their own work. Renderers change, and every studio has its own preferred toolchain, so quizzing someone about a specific renderer mostly measures familiarity with a product. Asking about retopology measures craft.
This article explains what I listen for, why the topic matters for production, and how candidates can prepare without memorizing trivia. The goal is to help artists present their process honestly, and to help interviewers ask questions that predict real collaboration rather than surface-level tool knowledge.
2 What this asset does in the game
In a shipping project, a character or prop asset is never just a picture. It is a runtime object that must deform, cast shadows, receive lighting, collide with the world, and survive being duplicated, instanced, and streamed. When I ask about retopology, I am really asking whether the candidate understands that the mesh they hand off becomes the skeleton of many later decisions. A prop with chaotic topology may look fine on a turntable, but the moment a rigger tries to add a hinge, a bend, or a destruction state, the problems surface.
For characters, the stakes are higher because deformation is continuous. Shoulders, elbows, knees, hips, and the face all need enough well-placed geometry to bend without collapsing or pinching, and they need it in the right orientation. A candidate who can describe how they plan loops around a shoulder before they start cutting is describing production awareness. The same logic applies to hard-surface props, where bevels, support loops, and clean planar regions determine whether the asset reads crisply at gameplay distance or turns into a smeared silhouette under a normal map.
The asset also has to serve animation, effects, and audio teams who attach sockets, particle emitters, and collision volumes to it. A mesh with predictable naming, sensible pivots, and organized material assignments makes all of that easier. So the interview question about retopology is really a question about whether this person builds assets that other disciplines can build on top of, which is the actual job.
3 Style and art direction
Style does not excuse topology. A stylized character with exaggerated proportions still needs edge flow that supports the poses the animation team will use, and a realistic character still needs efficient density rather than brute-force detail. The difference between the two lies in where the silhouette carries the read and how much the texture maps are expected to do. In a stylized project, large flat color regions and crisp shapes mean the mesh itself must hold the form, because there is less surface noise to hide behind. In a realistic project, the high-poly sculpt and baked maps carry much of the detail, so the low-poly mesh can be leaner and more deliberate.
When I ask about retopology, I am also listening for how the candidate adapts to a given art direction. Someone who has only ever built one style may struggle when the brief changes. I want to hear that they adjust density, loop placement, and UV strategy based on the target look and the camera distance the player will actually experience. A prop seen from a fixed isometric angle can be optimized differently from a weapon held close to the camera in first person, and a boss character seen from across an arena can be treated differently from a companion who stands beside the player.
Good answers reference the read: what needs to stay sharp, what can be simplified, and where the texture will carry the burden. That is art direction applied to geometry, and it is far more informative than a list of preferred brushes or render settings. Style is a set of constraints, and retopology is where those constraints become concrete.
4 The production pipeline
The pipeline I describe to candidates runs from concept and blockout through a high-poly sculpt, retopology, UV layout, baking, texturing, integration, and optimization. Each stage hands something to the next, and retopology sits at the hinge. It receives an approved form from the sculpt and delivers a mesh that texturing, rigging, and animation can all rely on. When a candidate can narrate that handoff clearly, I learn more about their professionalism than any render can show.
I ask them to walk me through a recent asset stage by stage. Where did they block out proportions, and how did they validate scale before committing to detail? Did they retopologize before or after finalizing the sculpt, and why? How did they plan UV seams so that texel density stayed consistent across the character, and how did they handle islands for parts that need to move independently? These questions have no single correct answer, but the reasoning reveals whether the artist thinks in terms of a pipeline or in terms of a single finished image.
I also ask about iteration. In production, art direction changes, a rigger requests a loop, or a texture artist needs a seam moved. An artist who can describe revising topology without rebuilding from scratch is describing a workflow that saves the team time. That is the practical value of the retopology question: it invites candidates to talk about collaboration, revision, and handoff, which are the parts of the job that never appear in a portfolio render but determine whether a project ships smoothly.
5 Technical constraints to respect
Every asset lives inside a budget. There is a triangle count, a texture memory allowance, a draw call target, and a set of platform limits that vary by project. Retopology is where those numbers become real, because density decisions made in the mesh determine how much the rest of the budget can absorb. I listen for whether a candidate can talk about density in terms of where it is needed rather than as a single global number, and whether they understand that a clean low-poly mesh with good maps usually beats a dense mesh with mediocre maps.
UVs matter just as much. Consistent texel density, sensible seam placement, and islands that respect how the asset deforms or how materials are assigned all affect the final quality. I ask how they handle overlapping UVs for mirrored parts, how they pack islands to avoid wasted space, and how they plan for multiple material sets without fragmenting the asset unnecessarily. These are not glamorous topics, but they are the difference between an asset that imports cleanly and one that requires rework late in production.
I also ask about baking. A candidate should be able to explain how they cage a mesh, how they handle hard edges and smoothing groups, and what they do when a bake produces artifacts. The answers show whether they understand that the low-poly mesh and the maps are a single system. Technical constraints are not obstacles to art; they are the frame that makes art shippable, and retopology is where an artist demonstrates respect for that frame.
6 Polishing for the engine
The final stage is integration, and it is where many otherwise strong portfolios fall apart. A mesh that looks beautiful in a viewer may import with flipped normals, incorrect scale, missing material assignments, or a pivot in the wrong place. When I ask about retopology, I am also probing whether the candidate thinks about the moment their asset enters a project. Do they check scale against a reference, verify that normals face outward, and confirm that smoothing is set correctly before export? Do they name objects and materials in a way that a teammate can navigate?
I ask how they handle level of detail. A character or prop may need several versions at different densities, and the way those versions are generated matters. A decimated mesh can lose the edge flow that makes deformation work, so a thoughtful artist plans the lower levels rather than letting an automated tool decide. I also ask about collision, sockets, and any metadata the project expects, because these are the details that separate an artist who ships from one who merely renders.
Finally, I ask what they do when something breaks in the project. The answer usually reveals whether they can debug methodically: check the export settings, inspect the normals, review the material assignments, and compare against a known-good asset. That habit of verification is what polishing for the engine really means. It is unglamorous, and it is exactly why the interview question is about retopology rather than renderers.
7 Frequently asked questions
Why not just ask about the renderer they use? Because renderer preference is largely a matter of exposure, and studios often use different tools. A candidate can learn a new renderer quickly, but habits around topology, UVs, and handoff take longer to build and are harder to fake. Asking about retopology gives me a signal that transfers across toolchains and projects, which makes it a better predictor of how someone will perform on a team.
Does this mean high-poly sculpting skill does not matter? Not at all. Sculpting drives the form, and strong form is essential. But a beautiful sculpt that cannot be converted into an efficient real-time mesh is only half the job. I want to see both, and I weight the retopology conversation heavily because it is the stage most candidates skip when preparing their portfolios. If someone can explain their topology choices clearly, I trust their sculpting more, not less.
How should a candidate prepare? Build one asset end to end and document the decisions, not just the result. Be ready to explain why loops sit where they sit, how UVs were planned, how the bake was caged, and how the asset was verified after import. Honest answers about mistakes and revisions are more convincing than a flawless story. The goal of the interview is not to catch anyone out; it is to find out whether this person can make assets that a team can actually use, and retopology is the clearest place to see that.