My Level Creation Process

I’ve talked a bit about the creation process I use when building levels before, but this time, I thought I’d do a complete walkthrough of the process. This is the same process I use for all of my level designs, and I’ve iterated on it with each new level I work on. I developed this process by studying current industry standards for workflow; talking with industry professionals, university instructors, and fellow level designers; and what I personally feel comfortable with. I have broken everything down into steps, and each step has an estimated time associated with it, which simply represents the amount of time I typically spend on that particular portion of the level. Below is the quick list of steps I follow, but to better understand my workflow and creation process, read on!


Step 1: Conceptualization

Step 2: Research & Pre-Production

Step 3: Asset List & (Possible) Creation

Step 4: Blocking In & Game Play Testing

Step 5: Asset Implementation & Scripting

Step 6: Populating & Polishing

Step 7: Final Product & Retrospective



Step 1: Conceptualization

Estimated Time: 1-3 hours

Like most things, each new level starts as a simple idea. Where this idea comes from depends on a lot of things, such as the time of day, where I am, something cool I saw – anything can serve as inspiration. Usually, my ideas for levels start with either a basis for the environment, or a game mechanic that I want to incorporate. Once I have an idea, I jot it, and all other ideas that go along with it, in a notebook. I also start sketching top-down maps of what I might want the level to look like, although nothing is set in stone at this point. I like to sketch different versions of the same level, which are mostly floor plans/top-down maps, but it’s good to sketch elevations/side-views, architecture styles, landmarks, and anything else that will help to translate the level into digital form. I start sketching in a sketchpad, then move onto graph paper, and recreate the final version of the level in Photoshop. The conceptual stage is spent exploring ideas for the level, and is usually where I establish a visual style.

Below is an image of some notes for an early level I did. This one was for a class assignment, where we simply had to create a 3D level with some sort of enemy to attack the player. I jotted down a few basic ideas to start with, which can be seen in the upper lefthand corner (made in UDK, very tech-like feel, dim lighting, grungy, space center). Then, I started sketching different layouts; what you see is just one that I wound up using, but there were many more of varying designs.

Level Notes and Map


Step 2: Research & Pre-Production

Estimated Time: 4-12 hours

After I have a solid idea of what I want the level to be, as well as a few sketches of what I want it to look like, I move on to fleshing everything out. I like to put a lot of research into each level, so that it can look and feel the way I want it to, and be an accurate portrayal of the environment I based it on. This includes finding reference and inspirational images, as well as researching specific topics, atmospheres, lighting, layouts, etc. If there is any part of the level or its design that could benefit from research, I look into it further. For my swamp level, I did a lot of research regarding the atmosphere and look of swamps, so that I could make it as accurate as possible. But for another level, which was a simple hotel room, I researched themes, such as confusion and distrust, so that I could incorporate that kind of symbolism into the design of the room. Architecture is another great topic to research, and it can even have a profound effect on the way your level is laid out.

At the same time that I begin researching, I start what will be my pre-production document. To start with, I simply take my handwritten notes and add them into a document. Then, I use my template for level pre-production to create different headings, which give the document structure and keep me on track as to what all I should have in the document. Next, I take those original notes and expand on them. I incorporate the research I’ve done into these sections, although I might have a separate document specifically for research, just to keep the pre-production from becoming too cluttered, especially since I probably won’t use everything that I looked up.

Here’s a quick outline of what my pre-production template contains:

  • High Concept
  • Features
  • Level Map and Layout
  • Objectives
  • Detailed Walkthrough (single-player maps)
  • Choke Points (multiplayer maps)
  • Controls/HUD
  • Visual Style(s)
  • Setting
  • Atmosphere
  • Sounds
  • Photo References

Those are the basic sections in my template, but depending on the level, I might take out or add in other sections, so long as the documentation serves its purpose and explains what the level is about. I personally like to write rather detailed documents, which range from six to twenty pages, depending on the complexity of the level.

By this time, I already have a good idea of what the level’s layout is, but I still like to sketch out a few other variants, just in case plan A doesn’t work out the way I want. For some levels, I try to think modularly, where the level is divided into sections that can be swapped out with others. I’ve found this kind of design to best for creating interior levels, although it definitely could work for exterior, if done right. The reason for doing this? Sometimes it can be difficult to tell how a level is going to flow on paper, and your level might not play the way you want it to when converted to three-dimensions. If I am unsure of how my level will work, I like to create a simple scale model of the level. These don’t need to be super fancy or anything, but they should be a relatively accurate rendition of the level in 3D. I’ve created scale models for levels out of stacks of Magic cards, random boxes lying around, and anything else I can find. Below is an image of a much more complex scale model I created, but this one is kind of an exception as to the amount of detail:

Detailed Scale Model


Step 3: Asset List & (Possible) Creation

Estimated Time: 2-3 hours (not including asset creation)

At this point, the pre-production is complete, and it’s almost time to open up an editor. Before I start working on a level, I like to make an asset list, which simply lists all of the assets I plan on using, or could use, for the level, along with where they are located or if they need to be created. Modeling isn’t exactly my forte, so I tend to stick to pre-made models, usually the ones that come with UDK. I do enjoy creating or finding new textures, though, and these are also added to my asset list. The asset list isn’t all inclusive, but it does give me a good idea of what I have to work with. After I’ve made this list, it’s usually time to find new resources, unless I plan on using pre-mades only. If I plan on making my own, that takes a big chunk of time out of my day, so that number is not included in the estimated time above.


Step 4: Blocking In & Game Play Testing

Estimated Time: 3-10 hours (depending on the number of iterations)

It’s finally time to start creating the electronic version of the level! I always start by blocking-in, and this is standard for the industry from what I’ve heard and read. The reason for doing this is to test the level to make sure it flows and works the way I want it to. I start by converting my map into a texture that I can import into UDK. Then, I create a giant, flat plane, and apply the map material, which I use as a guide for adding BSP brushes (or simple 3D shapes). Below is a picture of the “before and after” of blocking in the swamp level I created recently:

Blocking In Process

After the first pass of blocking in is done, I test out the level. The first play test is to see how the level’s design flows, and to make sure there aren’t any issues with the layout. If I come across a problem, I will make the necessary changes to correct it, or, if it’s a major problem, switch to another design entirely. This is an iterative process, and it’s important to make sure that the level functions properly before committing to the details. After I’ve got the basic level down, I move on to implementing some important game play features. These are usually simple things, like setting up the flags in a CTF map, or marking where key objectives might be. I only add in what is important for making sure the level will play the way I want, so no scripting or anything like that. For the swamp map, I blocked in where the trees would be and added in water volumes, so that I could test how players might move around the level. Play testing is crucial here, so do it often.


Step 5: Asset Implementation & Scripting

Estimated Time: 6-30+ hours (depending on the complexity and size of the level)

So I have a semi-functioning level and I’m happy with it, which means I can move on to adding details. The first pass at asset implementation is usually where I start, and this includes adding things like terrain, structural supports (floors, walls, ceilings, etc.), and anything else that is important to the general look and feel of the level. Again, I try not to get too detailed here, since that’s what the polishing phase is for, but it’s good to make it actually look like a playable level. Also, I cannot stress enough to save often. Level editors have a tendency to be a bit buggy, and it’s easy to lose hours of work, so save often and save several versions. Every time I make a major change to the level, or add in a bunch of assets, I always save it as a new version, just in case something happens and I need to return to a previous save.

With the first pass of asset implementation complete, I move on to scripting. Although some scripts may change after polishing, I like to add all of the functionality in at this point, since you can always tweak things later. The actual scripts vary per level, but a few general types that I’ve included in my levels are bot functionality, cut-scenes, object movement (doors opening/closing, boulder rolling towards player, etc.), light switches, sound effects, custom items and their effects, and a lot more. Depending on the number and complexity of the scripts you want to include, it could take a lot of time to complete this step. For my Inkblots level, I spent a large portion of the total 60-70 hours adding and refining Kismet sequences. Also note that every time you add in a new script, you need to test it to make sure it works properly, and because it probably won’t, be prepared to adjust things and try again.


Step 6: Populating & Polishing

Estimated Time: 10-30+ hours (depending on the size of the level and the number of items and details to include)

I should now have a fully playable level at this point, but it probably doesn’t look very good, so this next step is dedicated entirely to polishing it up. For me, the phrase “polishing” includes populating, lighting, post processing, adding details, adjusting scripts, and generally just making the level look its best. I try to always start with populating, since it’s easier to do everything else afterwards. Populating for an exterior level includes foliage, rocks, and any other objects you have in mind, while populating for an interior level means placing various objects, such as chairs, tables, beds, cabinets, crates, etc. This part is fairly time consuming, as you want to make sure that there’s some variance between objects. This is the most difficult for exterior levels, especially those with a lot of trees, since you don’t want all your trees to look or be facing the same way. That being said, populating a level will make it look more realistic and “livable.”

Next up are the details. This is usually where I go crazy with textures/materials, as well as adding smaller details, like a patch of ivy growing along a wall. While this part does take quite a bit of time, I find it to be rather enjoyable. Speaking of details, this includes adding in sound effects and ambient music, if you plan on including these. Then, I move onto lighting, which can be difficult to get to look right sometimes. I try to keep my lighting as realistic as possible, and the best way to achieve this is to refer to your reference images (that’s what they’re there for!). After lighting comes post processing and adjustments to scripts and anything else that might need adjusting. Again, the polishing phase is all about making your level look good, but that also includes tweaking things like trigger times, cinematics, firing speeds, etc.


Step 7: Final Product & Retrospective

Estimated Time: 1-3 hours

The level has been completely polished, and it appears to be done, so what else is there left to do? Well, first off, you need to make sure it works! I always do a full play through at the very end to make sure that everything looks and functions as I intended. I’ve come across some bad bugs or missed areas that I might not have noticed otherwise by doing this, so take the time and try to break your own level. You might just be making a particular level for a class or portfolio piece, but if you want to be a professional level designer, you have to start thinking like one. If a real player were to go through your level, would they get stuck anywhere or be able to do anything they shouldn’t? In addition to this, get other people to play through your level. Friends, family, random people you met on the internet – anyone can provide feedback and help you improve your work. Not everyone will be able to play your level, as I’ve found, so it also helps to take screenshots and videos. I take the feedback  I get and apply it, so even though the level may be complete, I’m still trying to improve it and thus, improve my skills. 

The last thing I do is to write a retrospective. This is actually something new that I’m trying, and I think it’s working well. I break down the level into several sections: what went right, what went wrong, what could be improved upon, lessons learned, and personal reflection. This is a great way to put my analysis skills to work, and by understanding what worked and what didn’t, I am growing stronger as a designer. Even if you never show anyone your retrospectives, it’s something I definitely recommend doing, since you can then take what you learned from creating that level and make something even better.


I hope that this has provided some insight to my level design and creation process! If you have any questions or comments, feel free to leave a comment or send me an email. Thanks for reading!

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: