Level Design

Sources

Explanations
Talks
References

Workflows

Pre-Production

  • Text-based planning .

    • A Portal 2 level is planned in text form.

  • Planning with Bubble Diagrams .

    • Just a different way to organize ideas, like organizing them in text.

  • 15 questions a level designer should ask about their level .

    1. Why is this interesting?

    2. Is it clear to players why this is interesting?

    3. What is this game really about?

    4. What kind of spaces work best for this game?

    5. Is this level/area bigger than it needs to be?

    6. Is my level needlessly flat?

    7. Am I just doing this because it’s what games usually do?

    8. Should I get feedback on this now?

    9. Is this really necessary?

    10. Is this how players will actually play it?

    11. Are the different areas and beats of this level distinct?

    12. Are there any interesting surprises?

    13. Is this too cliche?

    14. What’s the relation of this level to the other levels in the game?

    15. Is this level intrinsically interesting?

  • Another 15 questions a level designer should ask about their level .
    16. Does this level have the right amount of ideas in it?
    17. Is there a simpler way to do this?
    18. Could I make my objectives more narratively interesting?
    19. Are there any mechanics that are underused?
    20. Can I make my NPCs less generic?
    21. Can I make this level/game less thoughtlessly violent?
    22. Can we do that cutscene or story idea more interactively?
    23. What if I had to ship this level next week?
    24. How does my level come across in terms of gender representation?
    25. Will players of all skill levels have a good time?
    26. Is this just filler?
    27. What happens if the player just runs through the level?
    28. Is it important for this sequence to be challenging?
    29. Could I just cut this?
    30. Am I working too much, or too hard?

  • Jump height, jump distance, distance between points of interest .

Blockout

  • "Maximum information, with minimal effort".

  • Blockout and Affordances .

    • It's about level design in Uncharted Lost Legacy.

    • I didn't find the second part of the video very useful, since I already knew the concepts and have seen them presented better by BlenderGuru.

    • {1:21 -> 27:19}

      • The video gives good tips and explanations on how to create natural guiding lines for the player to make navigation smoother through the level.

      • It also explains the use of "blockmesh", which is basically the same as 'level design prototype blocks'.

  • Common problems in blockouts .

    • Too big.

    • Too symmetrical.

    • Too flat (non-vertical).

    • Too open.

    • Too empty.

    • Too linear.

    • Too samey (no variation).

    • Too generic (no uniqueness).

  • Stairs .

    • Interesting.

    • {1:46}

      • Talks about the ideal proportions for realism.

Action Blocks

  • Action Blocks .

    • *About:

      • About level design in Titanfall.

      • Discusses using Action Blocks to plan and produce ideas that are interesting and add to the gameplay.

      • The whole video shows the design process for the game's singleplayer levels while commenting on design problems and ideas that did not make it into the final game.

      • It's interesting to watch. It's somewhat relaxing to see an explanation that includes failures and development prototypes.

    • It explains how focusing on 'cool action movie scenes' had a negative impact on the game's development by not actually developing the game's mechanics. Action Blocks were proposed as a remedy. They are a kind of game-jam within the same game, made as cheaply as possible visually, aiming only to develop mechanics WITHOUT 'heroic cool moments'.

    • The timeline for an Action Block is 1–2 weeks per Action Block. They spent about 6 weeks on this approach.

    • The purpose is to inspire and share ideas. It's not necessary to choose any of the developed mechanics.

Modularity

  • Modularity, Granularity and the use of Kits .

    • *About:

      • It's about level design in Fallout 4.

      • Very focused on modularity and element granularity when building a 'kit'. Very interesting. Useful when planning large-scale games.

    • {16:16 -> 17:26}

      • Explanation about granularity of kit elements. This reduces redundancy when making changes while giving freedom for modification and easier terrain destruction.

    • {17:33 -> 19:16}

      • Discusses the problem of working with granularity and how to handle the increase of objects in a scene. The approach used was to create Prefabs (Godot scenes) that group granular objects inside the engine to reduce the number of loose objects in the level.

    • {19:16 -> 22:00}

      • Addresses the workflow involving the kit and how it communicates with other stages of level creation.

    • {22:00 -> 28:15}

      • Very interesting part of the talk about priority  when deciding which assets should be made first. The approach used is generic pieces -> variant pieces -> hero pieces . This comes from a level designer perspective rather than an environment artist, but it shows that the most tedious pieces to make are often the ones that most signal progress in game development.

    • {28:21 -> 31:20}

      • Interesting demonstration of the advantage of using kit granularity, showing different scenarios built using granular design.

    • {32:23 -> 34:59}

      • Use of kits that don't rely on snapping, e.g., construction kits, cosmetics, irregular things.

    • {44:31 -> 52:12}

      • More demonstrations of modular assets using granularity.

      • Shows some assets that use a 'socket' system allowing one socket to fit into another's 'hole'.

      • At the end it also shows the Prop kit used to create machines and so on, further illustrating the advantages of modular design even in non-regular situations.

Walkthroughs

Half Life 2
Quake

Concepts

Philosophies

  • !

  • Fun to navigate.

  • Don't rely on words / Tells what, but not how.

    • Let the player fill the gaps with creativity and exploration.

    • Give choices to the player, letting them explore and be creative.

    • Be creative in ways to tell the narrative. Don't give walls of text to be read.

  • Constantly Teaches / Is Surprising.

    • Example from Zelda, which turns the game into a huge tutorial: a new mechanic is introduced, many ways to use it are taught, then it is tested in a "boss"; sometimes after the boss the mechanic returns and surprises the player with another use that wasn't explored.

    • Creative re-use of mechanics.

    • Beware of being predictable. Disturb expectations.

  • Deliver the fantasy.

    • Adapt challenges based on atmosphere and the game's pitch. For example, make challenges easier when they conflict with narrative or the game's focus.

    • Let the fantasy and emotion guide you, then design the level.

  • Is efficient.

    • In short, rely on reusing content without making it feel repetitive.

    • Modular design, where mechanics return in familiar ways.

    • Bi-directional design, so that GOING to a level is very different from RETURNING from it, making it interesting and better than simple backtracking.

    • Create levels and mechanics that are meaningful.

  • LevelDesign: Design philosophy tips .

Composition

  • Affordances / Intentionality, World Building and Interactive Narrative .

    • {3:45}

      • Shows a slide with the video's premise: the connection between Presentation, Gameplay and Story.

    • {4:05 -> 24:42}

      • Affordances and Player Intentionality.

    • {24:42 -> 35:05}

      • World Building.

    • {35:07 -> 49:20}

      • Interactive Narrative.

  • Spatial Communication .

    • Great video covering level design fundamentals as the speaker navigates the created world.

  • Interior Design for level design .

    • Excellent video. Very dense and valuable.

    • The video ends at {47:30}.

    • The speaker is Dan John Cox.

  • Refuge Space and Prospect Space .

    • It's about level design in HyperLight Drifter and Sunset Overdrive.

    • Not a very practical video, but I found the concepts of 'refuge space' and 'prospect' interesting.

    • Recommends the book 'An Architectural Approach to Level Design - Christopher W. Totten'.

    • {11:55 -> 15:44}

      • Explanation and examples of 'Refuge space' and 'Prospect', as ways to convey safety vs danger.

  • Landmarks .

  • Landmarks with Denial and Reward .

  • Framing .

    • Talks about 'funnels' and 'choke-points' to "control" the direction the player will be looking.

Narrative

  • Pacing .

    • Discusses 'gameplay beats' and action regulation. Very interesting.

  • "If you don't make down-time in your game, people will seek down-time outside of your game".

  • Bait and Switch / Plot Twist .

Analyses

  • I go into a bit more detail in my analysis in my personal notes. This is just here for: I forgot.

Hollow Knight

  • Grub challenges ; I watched them all.

    • The challenges do not introduce new mechanics, so solving them often depends on the player's execution skill.

    • Some involve parkour challenges.

      • Some of these involve an extension of the current parkour challenge in the level.

    • Some require metroidvania unlocks to obtain.

    • Some are obtained almost automatically, so the challenge is finding the grub; for example, hidden behind breakable walls, sometimes requiring following the sound.

Destiny 2

  • Imbaru Machine.

    • The introduction of encounter mechanics is poorly done, so sometimes the encounter ends and I still don't know why the solution worked.

  • Dungeons and Raids.

    • After thinking about all raids, I didn't find any that present mechanics in an investigatory or interesting way.

      • "Wipe and try something new".

        • Many dungeons and raids are based on that. They often require cooperation and understanding allies' POV.

      • "Reduce the mechanic to the absolute simple and induce that it be discovered, without forcing wipes".

        • An example is the Prophecy, Duality, which remove all elements except the relevant ones and then test the player with a trivial challenge; usually some wipes are spent to understand the full nature of the mechanic, so it still ends up a bit like the method above.