Designing a CS:GO Level

Intro & Context:

In this Level Design blog, I’m going to talk about the creation of one of my earlier competitive CS:GO maps (created in Hammer). The map itself is called Isla de Piedra (excuse my broken attempt at naming the map in a foreign language) and is set somewhere in Ibiza, Spain. The “islas baleares” are a beautiful group of Islands and I had the pleasure of visiting them some years ago. I actually have very fond memories of the interplay in modern and traditional architecture as well as the lifestyle in that region.

That was also one of the central themes I was going for when designing this map, trying to combine a modern architectural style with what might have been there before. In short: Isla de Piedra is a tiny island where somebody decided to build a private, luxurious, fancy and somewhat villainesque estate on. And that’s most of the narrative context I thought of before starting to create the map. The other major piece was the idea of the mansion functioning as a disguise for arms deals and weapon smuggling affairs. I wanted to combine a peaceful, “holiday” location with a dangerous undertone.

So essentially what happens is the following: Counter-Terrorists are approaching the Island by boat, trying to quickly take control over the mansion and evidence that could be found there. Meanwhile, the Terrorist cell realizes that their hideout has been taken apart by law enforcement and they return to destroy all the remaining weapons and evidence at once to hide all their traces. Simple enough, right?

Reference & Aesthetics:

To start off, I gathered a bunch of reference images about luxurious buildings within a coastal setting. These six are the main ones I picked, focusing on the overall mood and architectural style that I wanted to go for. The main color palette I wanted to establish was white, grey, and brown with other smaller, less prominent pieces of color (dots of light red, blue and green scattered throughout, as needed). The color palette was also chosen in line with the idea of having a map that is quickly and easily readable in the context of competitive multiplayer design. My intention was that the outfits of the CT’s as well as the T’s pop out in front of walls & other parts of the environment.

As a next step, I decided to quickly prototype the first piece of visual aesthetic in the editor before even starting with a layout. By creating this small diorama scene while using all the official assets I could find that was fitting and already used in the game in different levels, I was able to still create a style / visual identity of my own. As I was not a professional 3D Artist it was important to me that I could reuse enough of the classic CS:GO models to create a feel for the environment and what I wanted to go for. I also played around with a bunch of custom textures in this step.

But – I didn’t want to spend too much time on this step as I was more focused on establishing a first general layout, blocking that out and getting a feel for the space I wanted to create in and what was then required for the environment art part resulting from that space.

Layout & Blockout

Looking at my visual prototype and the reference I gathered, I came up with a rough sketch for my level layout. I didn’t want to spend a lot of time in making the sketch look fancy or anything as I was aware that there were many iterations to all the details and cover placements necessary anyhow.

Iteration throughout the process was key to me. I wanted the freedom to explore different ways of solving issues like timings, line of sight, chokepoints, areas to use smokes & grenades as well as the bombsites themselves. I also knew that it would most likely change a bit after playtesting the level with a bunch of friends.

As you can see from the layout sketch, I chose to build upon the classic three-lanes structure of the established CSGO meta gameplay that is present in all the other official competitive maps as well. You have two spawns on each side of the playable space, while the attacking side spawns farther away from the objectives (two individual bombsites) than the defending side, giving the defenders enough time to set up their “line of defense” before the terrorists come knocking. There are also a bunch of rotational pathways between the main lanes to provide more tactical depth to the attackers and how they choose to hit a bombsite. Verticality is also present to a small degree if you look at the fake-staircase lines I used to highlight those areas.

Once I was satisfied with the layout sketch I wanted to test out the scale & proportions of what I had in mind and quickly imported the 2D concept as a texture to the editor and simply put it on a plane (i hope I’m not the only one using this lifehack :D). I placed some assets & entities (cars, character models, walls & boxes with the established dimensions that work in other maps) around the literal map of the level and scaled it up & down until it felt right. The whole sketch could now be used as a blueprint for the first block out I wanted to make.

Image: Flying around the first playable version of the map, checking out scaling and proportions.

I started out by focusing on the rooms and boundaries I had in mind and added all the essential walls that were needed to make the flow work for now. In this next image, you’ll see that there aren’t any vertical elements within the walkable space and that’s exactly what I wanted to do as the next step. For now, it was all about figuring out the dimensions and timings it took to move across different sections of the map. I wanted to get a feel for how big the space was, which I was setting up as a blockout.

As previously mentioned I then started to focus a bit more on the verticality in the first blockout. The structures of the mansion/estate were being set up as the playable space and it started to feel more and more like a traditional “level blockout” so to speak. As you will see a little bit later down the line, some areas will change significantly, but the overall structure of the island I was setting up remained intact.

The more I was adding structure and depth to the blockout, the more I was getting a feel for how the overall map could look like and play. Each time I was adding and changing elements of the blockout, I booted up the game and checked it out in-game to see how it felt there. I made adjustments according to what I thought was wrong or missing to the geometry of the level.

For the general layout, it was also important to me to support all variations of gunplay. There are parts where Sniper Rifles will excel and there are parts where medium to close range weapons such as SMGs and pistols will have an easier time. For example, the route below with the balcony in the next image is the “lane” I wanted to be destined for long-range combat. This is also the lane you chose if you want to immediately engage with other players, similar to the T-Spawn and DoubleDoors-CT line of sight in the center of de_DustII.

All other lanes in the map are more suited to other weapons than snipers and it takes a bit more time until you meet enemies. That does not mean that snipers become generally unviable in the map, said lane is just a place where they have a bigger advantage than other guns. As the financial economy in CSGO plays a huge role, I wanted to provide the attackers and defenders with different gameplay & gunplay opportunities throughout the map.

The more time I invested in the blockout, the more detailed it naturally got. I was tweaking different parts of the geometry and defined the spaces that were part of the gameplay area and which parts of the mansion were not gonna be accessible at all and only needed to function as “backdrop”/decoration. Also, some POIs & smaller “landmarks” were added to the level too, to give the players some orientation to where in the map they were.

I also now started to do more playtesting with actual other human players and started to figure out how the timings for meeting in the middle of the map as well as in front of the bombsites looked like. It was important to know where players would be able to get to in a certain amount of time, so I could design the chokepoints and bombsites accordingly and balance things out. Getting the timings right is one of the most essential parts of competitive map design as the competitive viability and balance of your map will instantly collapse if not tested vigorously.

It was also hugely important to me not to make the interior part of the buildings feel too claustrophobic because that’s always annoying for players when they cannot move and jump around the space how they’d like to. Having played 1000+ hours of competitive CSGO myself, I already know what you do as a player when moving around the map so I tried to support that way of traversal in this level (sidenote: there will be a video next, showcasing just that – look at me playing around with the knife and how I am constantly hopping around).

I tried to also do some first small environment passes on certain chunks of the map to see if it would fit with the scale I had blocked out and because I enjoyed doing that for a bit of the time. It enabled me to look at different parts of the map with fresh eyes (and not only from a purely design-oriented perspective but from an art standpoint as well).

Next, you can now find a video where I run around the “kinda final” (its never final until the map is done) blockout level for a couple of minutes. That way you can get a feel for the map and how it looks & feels like from down “in the trenches”. After the video, I’ll talk a little bit about each of the bombsites.

You can see in the video that I’m trying to show you the spatial compositions of each area of the map. Some areas you will notice will be “blocked off”, some (where I shoot the glass for example or the vents) become viable pathways for the attackers when you interact with them and so on. The orange doorways are markings of where I want “locked doors” to be, which seal off non-accessible rooms of the mansion (they’re bangable, of course!). I also showcase some high-risk high reward “flanking” routes/areas that require skillful jumps to take advantage of (which I fail to pull off, so it’s already balanced).

After a bunch of playtest sessions with friends and other CS:GO mates, I decided to readjust the spawn for the CT’s to be closer to a bombsite and added a bunch of additional cover opportunities to help guarding the area.

As a general rule of thumb in CS:GO level design, its always bad for the gameplay to have too many “line of sight”-angles at the same time / the player would have to systematically check every single one of them for enemy presence when trying to take control of a new area of the map. By trying to keep the area layouts as simple as I could I tried to avoid making the player look around too much and tried not to overwhelm him/her with choices. Taking control of a bombsite is always a very planned and systematic approach in the competitive environment. Teams plan their approach to each bombsite most of the time way before the round starts and the try to take control over an area as quickly as possible when the decision is made.

For both bombsites it was super important for me that defensive positions were not too overpowered and that attackers had the chance to block of the vision in certain hallways, stairs or spots of cover with smoke or flash grenades so that they were not eliminated from unfair angles and had opportunities to take control for a part of the map.

Also, the provided covers for the CTs should always be able to be combatted with Molotov cocktails from a T-POV, so that you can force the CTs to come out of their hiding spots and make them show themselves. It is still important, though, to not debuff the “flow off covers” (read: falling back from line of defense to next line of defense) too weak. If the defender has a great spot than making him fall back should be more punishing than defending a weaker spot where you would naturally have better cover if you would fall back.

Ultimately you also have to take into account the rotations between bombsites, speaking of CTs trying to help out other defending CTs at different parts of the map. Normally rotating takes about 12 to 15 seconds from one bomb site to the other so I tried to also design the lines of defense in a way that you could reach and support the other bomb site in that same timeframe.

It’s also interesting to note that once one of the bombsites is being captured by the attacking side, the roles switch around. Now that the CTs become the attacking force and Terrorists need to defend the bombsite, you have to account for that by also providing cover for rotating CTs in the connecting pathways as well as around the bombsites. The same applies to the defending Terrorists. The dynamic of your map will completely change when the gameplay switches from “Defending and holding the lines until attackers perform their attack on a bombsite” to “the Bombsites need to be retaken by the defenders and the attackers need to protect their planted bomb”. Your level and blockout need to account for both scenarios and dynamics that will ultimately be present in a round of 5vs5 competitive Counter Strike.

After the blockout was kinda set in stone, for now, I started to focus more on the interior parts of the gameplay, trying to give each room more of its own identity and a narrative context that made sense to me. As an example, I turned this room you’ll visit first when moving through one of the “lanes” into a lobby area.

Turning the Blockout into a real Place

After numerous playtests and designing the interior parts of the level, it became more and more set in stone from a gameplay perspective. I then started to make some art passes on certain areas. In this phase, I didn’t want to go too much into detail of how these areas could be designed because I was always changing some bits when I playtested the level, but I tried to give the overall map a visual art pass regardless. Always at the amount of detail that it wouldn’t take too much time to change and switch things around.

From this overview of the level, you can see that some areas start to feel like real places instead of in-editor blockouts. Looking at the right side of the last and this next image, you can see that I set myself up with a selection of textures I would re-use again so I wouldn’t have to open up the texture browser every time. Kinda like a “texture palette”. In these next picture, you can see how the level transforms more and more into a “real place”. For the most part, I reused the textured that were already in the game but I also added a lot of custom textures. Not custom models were used yet because I planned to replace them with custom ones later on. They were of course perfect already for what I was trying to accomplish with the level.

At some point, I realized that it would probably help the player orientation a lot if from one side of the island you could see the mainland. So there we have our “huge landmark” crossed off from the list as well.

Once I was happy with the outside part of the map, I started to turn the interior parts into something more lively as well. I chose to give most rooms their own individual color palette so that you would immediately know where in the map you are, just by looking at the colors. I didn’t do much of the lighting yet as I was planning on turning the map into a different lighting setting / time of day (late sunset).

I also tried to provide players with the opportunity to line up smokes, meaning they have pieces of the environment (for example the ceiling windows, which are breakable of course) they can line up with their crosshair to throw a smoke grenade across the map from a safe location and make it land perfectly and reliably to be effective in the process of an attack.

This already brings this Part of the blog to an end and I’ll leave you with some more images from the current state of the level. I hope you found this little blog at least a tiny bit insightful into how CSGO levels are being designed and what you need to think of to make it an enjoyable experience for competitive players. Stay tuned for a short summary.

If you want to get in touch you can follow me on twitter (@sdgiesselmann) or send me a direct message almost everywhere.


When you want to design your own CSGO map, try to keep the following aspects of CSGO level design in mind:

  • CSGO Gameplay Conventions: Two Spawns, Two Bombsites, Two Teams, Attacker & Defender gameplay
  • Provide opportunities for utility usage (smoke, Molotov etc.)
  • Line of Sight in the spaces you create
  • Chokepoints, Line of Defense
  • Timings, Player Rotations
  • Readability (Colors you use in the environment)
  • Round Dynamics: defenders become attackers, Retakes will happen
  • Account for weapon variety, Think of financial Economy
  • The feeling of the Space, Scale, Dimensions
  • Design for the Player Controller & Traversal (Fun to navigate)
  • Narrative Context
  • Orientation & Player Guidance (Landmarks, Colors)
  • Try to avoid doors & ladders, they’re generally not fun to use if you need to rely on them in the heat of battle (otherwise they’re fun!)