The PCG framework in UE5.7 going production-ready has reignited an old debate: should levels be procedurally generated or hand-crafted? The answer, as with most game development questions, is "it depends" — but we can be more specific than that.

The Case for Hand-Crafted Levels

Hand-crafted levels have strengths that procedural generation can't replicate:

Narrative precision. A scripted set piece — the moment you open a door and see the dragon for the first time — requires exact control over camera angles, lighting, prop placement, and timing. Procedural systems can't author narrative moments.

Curated difficulty curves. The best level designers tune challenge progression by hand. The gap between two platforms, the placement of a healing item before a boss, the sight line that lets you spot an ambush — these details require human judgment about player experience.

Unique identity. Players remember hand-crafted spaces. The Hub in Dark Souls, Zelda's Temple of Time, the Citadel in Mass Effect — each is memorable because a designer made intentional choices about every corner.

Quality floor. Hand-crafted levels have a guaranteed minimum quality. A designer sees every part of the level and ensures it works. Procedural systems can produce broken, ugly, or unplayable results that require validation.

The Case for Procedural Generation

Procedural generation solves problems that hand-crafting can't:

Scale. A 100 km² open world can't be hand-crafted by a small team in a reasonable timeframe. Procedural systems make large worlds feasible for indie developers.

Replayability. Roguelikes, survival games, and exploration games benefit from layouts the player hasn't memorized. Procedural generation makes each run feel different.

Development speed. Procedural rules can populate an environment in seconds. The same work would take an artist days or weeks by hand.

Consistency at scale. A well-tuned procedural system applies rules uniformly across a large space. Hand-placed environments can have inconsistent quality, especially when multiple artists work on different sections.

Iteration speed. Changing a procedural rule affects the entire world. Changing a hand-crafted placement affects one spot. For broad art direction changes, procedural wins.

The Spectrum, Not the Binary

The debate frames this as either/or. In practice, most shipped games use both. The question is where each approach gets applied.

Fully Hand-Crafted

Best for: narrative games, puzzle games, fighting games, racing games (tracks), competitive multiplayer maps.

These games need exact control over every element. Procedural generation adds variance that hurts the core experience.

Fully Procedural

Best for: roguelikes, infinite runners, some survival games, sandbox games.

These games need variance as a core mechanic. A hand-crafted roguelike defeats its own purpose.

The Hybrid Approach (Most Games)

Best for: open world games, RPGs, adventure games, simulation games.

The hybrid approach uses hand-crafted design for high-value areas and procedural generation for large-scale fill:

• Hand-crafted: towns, dungeons, quest locations, narrative set pieces, boss arenas
• Procedural: wilderness vegetation, rock placement, ground cover, ambient wildlife, distant terrain detail

This is where the Procedural Placement Tool fits. It's not designed to generate levels — it's designed to populate the spaces between your hand-crafted content.

How to Decide: A Framework

For each area in your game, ask three questions:

1. Does the player interact with it mechanically?

If yes — if the player climbs it, fights in it, solves a puzzle on it, or navigates by it — hand-craft it. Mechanical interaction requires precise tuning.

If no — if it's background vegetation, distant mountains, or ambient decoration — use procedural placement. The player experiences it visually, not mechanically.

2. Does it need to be the same every playthrough?

If yes — if a quest requires a specific cave layout, or a story moment depends on a vista being visible from a specific spot — hand-craft it. Procedural generation can't guarantee specific results.

If no — if the exact arrangement of trees in a forest doesn't affect gameplay — procedural is fine. Variety can actually be a benefit.

3. How much of it is there?

If it's a single room or a small area — hand-craft it. The time cost is manageable, and the quality benefit is high.

If it's kilometers of terrain — use procedural generation. No team can hand-place every tree in a forest that stretches to the horizon.

Practical Examples

Open World RPG

• Hand-crafted: City layouts, dungeon interiors, quest-specific locations, camp sites, ruins
• Procedural: Forest vegetation, rock scatter, grass and ground cover, wildflower placement, distant terrain detail
• Tool: Procedural Placement Tool for vegetation and environmental scatter

Survival Crafting Game

• Hand-crafted: Building blueprints, crafting recipes, unique landmark structures
• Procedural: Resource distribution (ore veins, tree clusters, animal spawns), terrain features, biome transitions
• Hybrid: Use procedural rules for resource placement but hand-place key landmarks as navigation anchors

Horror Game

• Hand-crafted: Every interior, scare trigger placement, lighting, sound cue positions, narrative elements
• Procedural: Exterior vegetation around the building, ambient detail objects, fog distribution
• Note: Horror requires precise atmospheric control — use procedural only where the player's attention won't be focused

Roguelike

• Hand-crafted: Room templates, encounter designs, boss arenas, item pools
• Procedural: Room arrangement, corridor connections, item placement within rooms, enemy composition
• Hybrid: Hand-craft the building blocks, procedurally assemble them

Common Mistakes

Using procedural generation for areas the player examines closely. A procedurally placed tree at the edge of a cliff looks fine from 200 meters away. Up close, it might be floating, clipping into rock, or placed at an impossible angle. Areas the player can walk up to need human review.

Over-relying on hand-crafting in large spaces. A solo developer hand-placing every rock in a 10 km² map will either burn out or ship a barren world. Use procedural placement for volume and hand-craft the focal points.

No validation on procedural output. Procedural systems can place actors inside geometry, on impossible slopes, or in gameplay-breaking positions. Always include validation rules — slope limits, collision checks, exclusion zones near paths and buildings.

Ignoring the uncanny valley of nature. Random scatter looks artificial. Real environments follow ecological rules. Your procedural system needs rules for slope, altitude, clustering, and species distribution to produce believable results. We covered this in depth in our biome design post.

The Right Tool for Each Job

Procedural generation and hand-crafting are tools, not philosophies. Use each where it produces the best result for the least effort:

• Hand-craft what the player interacts with
• Generate what the player looks at
• Validate everything procedural systems produce
• Iterate faster by using procedural rules for broad changes

The Procedural Placement Tool is built for the hybrid approach — rule-based scatter for vegetation and environmental detail, with exclusion zones and density painting for hand-crafted refinement.

Browse the documentation for setup guides and practical examples.