Creating Grail: Colorful Mini-Golf Park, Winner of Counter-Strike 2 Map Contest

Grail

Grail is a community-made 5v5 bomb defusal map currently available to play in Counter-Strike 2 and the 1st place winner of the Mapcore Big Adventures Mapping Contest.

The map was created by Bartlomiej “Lizard” Guzek, Radu Tanasie, and Florian “Flo” Wagner as part of the aforementioned contest that ran for an entire year, starting in April 2024.

Contest

The Big Adventures Contest was organized for Counter-Strike 2 by Mapcore and FACEIT in which contestants were tasked with creating original 5v5 bomb defusal or hostage-rescue maps.

The contest was split into 3 stages, each lasting several months. At the end of each stage, the maps would get judged by Mapcore veteran judges such as FMPONE, TopHattWaffle and Helder Pinto, as well as CS2 Guest Experts, including the likes of Launders, Anders Blume, James Bardolph and a few other notable names, further narrowing down on the number of maps until only 10 were left to compete in the final 3rd stage. The judging categories for the maps were: fun factor, originality, visual presentation, and overall polish.

The top 4 winning maps competed for a monetary prize pool of $15,000 ($7,500 for first place; $4,500 for second place; $2,000 for third place; $1,000 for fourth place). 

On the 4th of May 2025, it was announced that the winners were: Grail – 1st place, Alpine – 2nd place, Palacio – 3rd place, and Stronghold – 4th place.

Team

Bartlomiej “Lizard” Guzek – senior level designer at Bambaa, previously at Techland, Counter-Strike: Global Offensive community contributor.

Lizard: On one autumn day in 2008, when I was 14 years old, a random Counter-Strike 1.6 community map sparked my interest in map-making. 11 years later, one of my maps got added to the official Counter-Strike: Global Offensive map pool. The same year, I landed my dream job at Techland as a junior level designer, where I had the chance to work with some of the most talented people I have met while shipping Dying Light 2: Stay Human. And now, here I am, 6 years later, writing this article along with my friends Radu and Florian about our latest Counter-Strike 2 project, Grail.

Radu Tanasie – level designer at Keen Games, previously at Crytek, Techland, and Ubisoft.

Radu: I'm a designer with a passion for creating things ever since I was a kid. Over the years, I’ve had the pleasure of meeting a lot of like-minded people and working across 3 different countries at 4 studios. I like games with a strong sense of atmosphere, and my all-time favourite game series is Half-Life. In my free time, I either build levels, watch Korean dramas with my wife, or try to stop one of our cats from eating everything.

Florian “Flo” Wagner – 3D environment artist at Running With Scissors.

Flo: German by passport, artist by trade. My environment art journey began in 2021. After graduating in 2023, I dove deeper into the subject, learning from super-talented people in the industry. Since then, I’ve finished a Venice-inspired environment, Venice Map, and after that began working on Grail while also starting to work at Running With Scissors and some other freelance projects. When I’m not crafting levels, I’m probably deep into Anno 1800, Assassin’s Creed, Life Is Strange, Battlefield, or playing games with my girlfriend. 

Theme

Lizard: As with every mapping contest, starting with a solid idea for a map theme can significantly improve your chances of creating something truly special and memorable. Funnily enough, Radu and I had roughly the same idea for a contest map. My initial idea was a stylized paintball arena, and Radu’s was a mini-golf park.

Radu: That was the point where it started to make sense for both of us to team up for this project. By ourselves, we would have probably delivered on something cool, but a bit more scaled back. Together, we gained a lot of confidence that we could create a truly unique map.

Lizard: We quickly decided to go with the mini golf idea since in order to sell a paintball arena you need a lot of brightly colored paint all over the place, which, in our opinions, could lead to visibility issues.

Radu: The mini golf idea was something that had been on my mind for a while by that point, and the contest proved to be a good opportunity to use it.

Lizard: We both wanted to try to do something stylized. After all, the theme of the contest was “Big Adventures”, so we decided to take on the challenge of delivering something a bit crazier than your typical Counter-Strike map.

Radu: I think we both had some prior interest in this type of make-belief setting, probably planted in our minds from playing Portal back in the day. Personally, I was more interested in the fake aspect of the location and being able to see between the seams. Lizard had the idea of going a step further to stylize it, and I thought it could be an interesting idea to explore.

Lizard: There were a few ways to go about that, but we knew in order to make the stylized part of the project work, we needed to go with something like a movie studio set (which had been done already), amusement park, water park, or something that generally has fake scenery. We knew from the start that going with a mini golf park would allow us to push ideas that normally would feel strange or out of place – especially for a game like Counter-Strike 2.

Scale, Sub-Themes & Art Style

Lizard: When thinking about the overall ideas and how we could push the stylized factor of the project, I wanted to focus on a few pillars that became the backbone of the project: scale – subthemes – art style. 

Scale

Lizard: Recreating a real-life mini golf course wasn’t a viable option, especially for a game like Counter-Strike, since this type of place is basically a big open area with some objects that, in the best-case scenario, might work as half-body cover. We needed something that could break sightlines, create significant negative space, and provide enough contrast between the different map areas.

After doing some research, we knew that we have to go bigger and more epic. So the direction we set was to go with something that players could feel immersed in and be a part of some kind of background storytelling.

Disneyland was a very obvious reference. It's a lot bigger, has a variety of subthemes, a lot of contrast, and some negative space. What was most important for us was the fact that once you are in, you barely remember that it’s a “fake” place. The only concern we had at this point was a believability aspect of it. Why would someone build such a huge and cool-looking place just to play mini golf?

Radu: It didn’t take too long before we threw the believability aspect of it out the window. My first sketch was a lot more grounded, but once we started to test ideas and block out the spaces, the scale of it quickly ballooned out of proportion. But we didn’t see a problem with that. We were more focused on building something that we considered to be really epic.

Subthemes

Lizard: Something like Disneyland is still pretty vague and doesn’t say much. There is a lot of room to fill in with your imagination. What would be your typical path when going to such a place? What would be your points of interest? Would there be a story to live through while traversing this place? What areas could you visit in such an arena?

At this point, I decided to do a little brainstorming on subthemes. I wanted to make sure that each area of the map had its own identity so that players could easily create a mental map of it. Communication in games like Counter-Strike is crucial. Not having to think hard about how to explain in which area of the map you saw your enemy can make a huge difference.

Together, Radu and I generated a good list of subthemes and areas that we could have on the map that felt unique and, most importantly, allowed us to create different gameplay scenarios for players. You want to support every type of Counter-Strike player, play styles, and player tools with your design, be it long range sniper battles or close quarter combat with shotguns.

We then decided to categorize these areas by their possible purpose because some of them are a better fit for something like a bombsite or spawn area.

Art Style

Lizard: All of these ideas could work independently without having any stylized elements to them, but we wanted to challenge ourselves by having a stylized aspect to the map. I wanted the map to not only have stylized materials but also stylized silhouettes and exaggerated proportions.

Radu: Although the more granular details of the art style come into play later on, it’s important to consider some things early on, even for the blockout. That way, you won’t be surprised negatively down the line. For us, that meant working from the start with slightly stylized proportions so that we could get a good impression of the place.

Lizard: I knew right away that this would be the trickiest part of the project. 

Firstly, it had never been done on such a scale in Counter-Strike, so there was no proof of concept that we could look at and learn from besides other games. The closest stylized reference we had in the Source 2 engine is the new Valve game, Deadlock, which was not exactly something we were looking for.

To balance this, we wanted to make sure there is a certain percentage of the environment that is really grounded and looks like your typical Counter-Strike map. That would give players some contrast and create “anchors” to help with navigation. I think we aimed for an 80/20 ratio when it comes to stylized and non-stylized areas.

Radu: It was important for us right from the start to design the layout with these things in mind. We looked at the big picture and tried to distribute these behind the scenes in key areas to achieve a balance between navigating a fantasy place and clashing with reality.

Lizard: Secondly, it was obvious to us that we are stepping on thin ice when it comes to player reception. Counter-Strike players have a rather vocal community when it comes to unorthodox projects. Grail could be either hit or miss. Nothing in between. 

Radu: Despite that, we moved forward with our crazy ideas and hoped for the best. If we were going to stand out from other map creators, we would have to fully embrace the weirdness and lean a lot more into vibrant colors.

Lizard: Thirdly, we were operating on a tight deadline, considering the fact that both of us, and later Flo too, work full-time in the industry, so there was little room for error.

Initial Blockout

Radu: After we had a good idea for the layout and areas distribution, we started to simultaneously block them out in the editor. Lizard started with the throne room (B site) and an adjacent tower. I started with the middle area of the map and the castle exterior. 

I wanted to quickly get a good idea of what the compositions could look like and see how viable a long middle lane would be. I thought the castle in the background, with the bridge and statue in front of the entrance, framed by the houses on both sides, created a strong composition, similar to something like Disneyland.

Lizard: One interesting bit about the map design is that we tried to tell the story through gameplay as well. Since the defending team's spawn point is in the castle and the attacking team starts outside of the city walls, I wanted to try to invoke certain feelings for both teams. So the defenders should feel like they are indeed defending the castle from the attacking team that is trying to storm it.

I thought this was a really interesting challenge to tackle with map design, and one way to achieve it was by putting the defending team's spawn point directly on the bombsite located in the throne room. 

Normally in a bomb defusal scenario the defending team spawn point is located somewhere in between both bombsites so that players first have to reach the bombsite leaving them no time to push further.

Radu: Of course, there are exceptions to that rule. Overpass is one such example, where the defending team spawns on one of the bombsites. But we didn’t want to just copy that map and call it a day, so we pushed forward with the idea of creating the feeling of a siege.

Lizard: On Grail, since the B bombsite also acts as a spawn point, players have a lot more time to push out of the typical defending zone on bombsite, in effect shifting the map control zone outside of the bombsite itself. This creates the feeling of a “frontline” that they have to hold in order to defend the areas behind them (including the bombsite).

I remember one of the playtests where we received the feedback from one of the players that they felt like the map was designed in such a way that the only “default” option for them to defend the bombsites is to play near the main castle walls since they provide the most protection. We were very happy to hear that because it validated our design choices and showed us that we created a fun flavor that players perceive unconsciously.

Radu: After blocking out the middle part of the map, I moved to the port (A site) and also began to slowly flesh out an idea for the attackers' spawn point, as well as the background area directly visible from their side. Lizard continued to tackle the throne room (B site) which was proving to be trickier to design than anticipated.

At this point, everything was pretty rough. During the initial blockout, you don’t want to spend too much time fiddling with details. The most important thing at this phase is getting some basic shapes in to get a feeling for the environment that you’re trying to build and narrow down the scale of the place. 

Once we had that first rough blockout, we went over all the areas, reviewed them, and iterated based on the feedback that we gave each other. Most times, we would spend at least a few hours in Discord talking about ideas and what could be improved. We often joked that our calls started to resemble the typical review session in the triple-A industry.

A Site Blockout

Radu: After the initial blockout of the areas was done, I returned to A site to flesh it out and give it a more strong identity. There wasn’t anything special about it visually, so I knew it needed something unusual to make it stand out from a typical port.

One idea I had during the initial pitch was to have giant squid tentacles that players would have to get close to for a memorable wow factor. I remember at the time, Lizard was worried the tentacles could be perceived as disgusting by some players due to their fishy nature. 

On top of that, we weren’t sure how feasible it would be to deliver these to a good enough quality. So right from the start, this idea didn’t seem too likely to make it through.

During one of our calls early in the initial blockout process, Lizard did a quick test using the spline tool available in Source 2 to test if we could easily create a tentacle in the engine. The initial result looked a bit crude, but promising, so I took it over from there to try to create some cool looking tentacles.

As I continued to experiment with the tentacles, I tried to give them as much character as I could and convey a sense of motion, no matter the viewing angle. If you look closely, you will see the tentacles are always gripping or twisting towards something. 

This also provided a good opportunity to move the lighthouse that Lizard initially blocked, directly next to the port area, and have a tentacle go around it as if it were trying to pull it down underwater, which ended up being a key image of the map.

After a few more days and some rounds of additional feedback, I managed to get these compositions for the port area. Most of these remained pretty close to the final version. A final addition to this came in the form of a squid head that Lizard blocked out in the editor.

Once the tentacles were in, we started to see the potential this theme had to bring something truly memorable in terms of CS maps. 

Design-wise, this site plays as you would expect from a classic Counter-Strike bombsite. The shape of it is pretty straightforward, easy to read, with clearly defined entrances and cover islands. Besides the tentacles on the site that also serve as cover, the geometry is angular compared to B site.

During the playtests, it was received quite favourably, and people appreciated the contrast between the sites. I think that’s one of the things we nailed really well.

B Site Blockout

Lizard: Apart from working on the overall design and art direction of the map together with Radu I was also responsible for some dirty editor work around the B bombsite part of the map. 

Once we had the middle and A bombsite area laid down, I was trying to find a good balance for the other part of the map. Right away the idea was to have bombsite B in the throne room in order to build on the idea of having a siege and clear “frontline” along the castle walls.

I started with a rectangular room shape to try to get the scale right, and right away, I noticed a few issues with this design.

Firstly, the long rectangular shape creates long sightlines along the room walls that somehow had to be broken up. 

Secondly, in order to make the room believable, it required to be a lot bigger than your typical Counter-Strike bombsite.

Thirdly, I was trying to create a bombsite that didn’t rely on crates sprinkled everywhere like in every other map, so at this point, I felt like it would be really hard to achieve with the issues I listed above.

Radu: I remember that for a while, bombsite B just wasn’t coming together. We were 2 playtests in, and people were avoiding heading to that site.

Lizard: So after those playtests, I decided to change direction and go for a circular throne room.

This design eliminates the problem of long sightlines along the edges of the room, creating a better flow because almost all entry points are directed towards the center of the room.

It also allowed me to be creative with the cover. I could use the big pillars as a full-body cover and golden piles for a half-body cover. The circular design also felt a lot more monumental and allowed for easier and better art execution.

Art Pass

Color Palette

Lizard: When it comes to the color palette, initially, we had two ideas. Mine was more grounded. It looked a lot more like your typical medieval castle made out of sandstone with a very conservative approach to bright colors.

Radu: I wanted to go for a more Disneyland presentation, featuring brighter and more vibrant colors, in order to capture the imagined fantasy of the place.

Lizard: Both ideas had their pros and cons. Mine was easier on the eyes but lacked the “oompf” and didn’t help with highlighting the important elements in the map.

I felt like Radu’s idea, while very bold and colorful, had a few issues too. For me, it started to look more like a bouncy castle type of structure with a lot of bright colors rather than something you would expect from a real-life place with such a scale, and that also plays well as a multiplayer map.

Radu: This proved to be quite a divisive topic, with both of us having strong opinions about our own vision, and would come up several times through development.

Lizard: Initially, we ended up in a small deadlock for a few days until we finally started to work towards a middle ground. We highlighted what works and what doesn't in each other's ideas, took the best out of them, and combined them into one. 

This way, we created a version in which we still had nice bright colors, good contrast in places where we needed it, good player readability and a lot of variation when it comes to color values. This version helped to bring attention to the important elements of the map such as main castle, tentacles or the lighthouse while also avoiding being overwhelming in the places where player visibility was the most important thing.

Beauty Corner

Lizard: Every time I feel like the greybox is going in the right direction and won’t change dramatically, I start to work on a small beauty corner to set the art direction in motion. 

I usually pick an area of the map that supports a good variety of materials and assets. Then I will try to art it up using stock assets. This is a super important part of the process because it helps the project in a lot of ways by:

• showing us how much custom content will be needed,
• acting as a test scene for the custom assets that will come later down the line,
• being a quality benchmark for the rest of the map,
• establishing a visual style and finding potential issues early on.

In Grail, I picked the middle area of the map for this phase because I felt that it had a good variety of structures such as the houses, castle, parts of the city walls, and backrooms.

Having this part of the map more or less arted not only helped us with the previous points, but it also allowed us to look at it from the player’s perspective versus looking at individual assets placed in a zoo map.

Materials

Lizard: Our initial 3D artist, Michael Jako, did a small test before leaving the project, which was a small house using realistic assets. This showed us that having a stylized silhouette with realistic looking assets won't really cut it. So we felt a bit stronger in our initial vision to go for fully stylized materials while trying to maintain a realistic approach to the smaller details.

So for example wood should look stylized but when you get closer and see how light is reflected off of it should suggest that the wood is in reality made out of some kind of plastic which made sense for us since we thought that this would be a way to do fake wood in a stylized amusement park.

Radu: When Michael dropped out, we found ourselves in a bit of a pickle. We had all these ideas and ambitions, but nobody on the art side to help us pull it off. This was happening 5 months in, right before we were supposed to start the art pass for the map, so we felt like we urgently needed to find someone.

Luckily, it didn’t take long before we found Flo. One of our friends, Thiago Klafke, recommended to us one of his mentees, Florian Wagner, who was supposedly the hardest-working on his channel. Flo, at the time, had finished his Venice environment a month or so prior, so he was already on our radar. Both Lizard and I were following him on social media before approaching him. For us, it was clear that he was motivated, but there was still the question of whether he would be able to get into stylized art.

Flo: When I joined the team, Lizard and Radu already had a strong opinion of the stylization of almost all props and buildings. 

In the beginning, I was skeptical, as I knew that this had never been done before in Counter-Strike and was a bold decision to make, considering this was my first time working on a CS map. But I trusted their experience and began working on some of my first stylized materials.

Lizard: The way we approached this was pretty simple, from my point of view. I took a bunch of stock assets and sent them to Flo so that he could try to remake them as stylized versions.

Flo: This was my first time doing stylized art. It turned out to be a lot of fun, you have more freedom in going wild with colors and shapes and use your imagination to hint at details.

After I completed the first brick and wood materials, together with trim sheet versions for each, Lizard could imagine the place a lot better and put together small art scenes in the map to test the art direction.

From the beginning on, Lizard emphasized a lot on exaggerating shapes and proportions while modeling which is typical for stylized art, but I had to get used to it because it felt unnatural to my thinking while doing models.

Radu: Of course, we didn’t get the final look right from the start. In the second half of the contest, based on feedback we got from the judges and players, Flo had to go through all the materials and make adjustments. People were saying that the materials needed more micro details up close.

Lizard: Not everyone was a fan of the stylized look. Some people liked it, while others felt that there was a middle ground to achieve. We took that feedback and tried to act on it.

We quickly discovered while tuning down the stylized aspects of the assets that there was a very thin line between a map that was clearly stylized on purpose and one that was stylized without proper execution. The map kind of fell into some type of uncanny valley where stylized silhouettes and proportions looked really bad with more realistic looking assets.

So we all felt that we either go fully stylized and utilize everything that this style has to offer or we try to play it safe and fall somewhere in the middle without reaching the full potential of either style

Flo: But we still wanted the map to feel like an actual fake park inside the real world rather than inside a fantasy universe. I had to make several iterations to achieve a more realistic and grounded look while still keeping the appeal of the stylization. I used a lot of references from actual theme parks to see what fake stylized props look like in real life.

What that meant for the materials is that I added a lot more grime and dust building up in corners and crevices, desaturated stuff because of sun influence and bleached colors on edges because of faster wear down of the paint. Also, using photo textures to blend inside Substance 3D Designer turned out to be a very important part of it.

Substance 3D Designer

Flo: Grail uses around 30 tiling textures, half of them being trim sheets. A nice thing about pretty much all CS2 maps is their consistent use of one master trim sheet layout. That means you can use most of the existing CS2 trim textures, apply them to your trims and they work perfectly. That also makes things for me a lot easier, as I could use the same template for all trim sheets in Substance 3D Designer. 

For nearly all tiling textures, I could use variations of one master material graph I did in the beginning. I found some of the techniques I used to make that graph on the awesome YouTube channel 3dEx.

In the beginning, the graph contains the option to select either a custom base shape or the trim sheet mask layout. This is because I could use a slightly altered version of this master material graph for all trim sheets. More about that later on. 

After that, you can select the option to add cracks using my custom cracks node. I tried to outsource a lot of the different components to keep the graph cleaner. My stone noise generator on the top is also mostly outsourced and can be replaced by a wood noise generator. 

I’m using a simple edge chipping technique that works with the Flood Fill to Gradient node in a specified direction, then multiplying a pattern onto it and using the Levels node to mask out the edge parts of the shape to pure white. Depending on the needs, this can be duplicated to include more different angles.

For the Albedo map, I combined different elements extracted from my height map, which include Overall Color Pattern, Curvature, Color Variations, Dirt, Dust, Broader Highlights, and Pebbles. This creates a rather stylized Albedo map, so I laid a photo noise on top of it to ground it more into CS2’s look and feel. 

To make the trim sheets, I first needed an ID map of the master trim sheet, which I made in Photoshop. This enabled me to mask out the different trim shapes inside Substance 3D Designer using the Material Selector node. 

For each masked-out trim, I could use an instance of the master graph and select the custom shape input. Then I just have to blend every texture of every instance with each other. Beware, this can get quite CPU heavy with 19 material instances updating at the same time. The benefits of this approach are a complete, independent control on each trim and an easy update of the whole trim sheet. 

A Bombsite Art Pass

Tentacles

Flo: The tentacles were a special kind of prop. They’re very prominent in the map, so they needed to pop and sell the look well. We wanted them to look very organic and transport the angry nature of the octopus onto them. You can see this in the way it breaks the light tower into 2 pieces and breaches through the restaurant.

Radu: During the art pass, we revised some of the tentacles in order to better fit the updated composition and introduce some final ideas. 

Some of these new additions included updating the tentacle gripping the restaurant to hold a giant golf ball, framing an area of the background with a rounded tentacle, adjusting the grip on the lighthouse (including the structure itself to be slightly bent) to convey a crushing force on it, a tentacle smashing through the roof of the dungeon area and an additional tentacle extending towards the castle from the bombsite.

By the end of it, the tentacles had proven to be a big group effort in order to get them to the final version.

Flo: Lizard and Radu initially blocked out the tentacles in Source 2 with the help of a spline. We made several iterations on how they should finally look before I came up with a pipeline on how to make them. Frankly speaking, the pipeline could be improved and made simpler, but here’s how I did it in the end.

First of all, we exported a very thin mesh version of the tentacle as an FBX out of Source 2, as I couldn’t find a way to export the spline directly. This was done so that after I imported it into Maya, I could select an edge loop along the tentacle that would very closely resemble the initial spline in Source 2, convert that back into a curve and export it as an .abc file.

Then I switched over to Blender, imported the curve, and used a modified version of an awesome Geometry Nodes setup by Retroshaper to create the tentacle from the curve.

Now in ZBrush, I made a low-poly version of the tentacle using Decimation Master. For me, that’s the quickest and easiest way to low-poly static props. The polycount depended highly on how close you’ll get to the tentacle and if it’s directly positioned on a bombsite (which needed higher accuracy in its shape) or on a spot that you didn’t fight much on. So the polycount fluctuated somewhere between 10K to 60K per tentacle. This may seem a lot, but it was totally in range without any noticeable performance issues. 

In Maya, I imported the low-poly version and cleaned up the topology. Unfortunately, this has to be done mostly manually and can be quite tedious. There are always verts breaking through the mesh, lying nearly over each other, or creating noticeable spikes. You want to clean them up before proceeding to UVs.

All tentacles use exactly one 2K texture set on a single tile. The texel density isn’t always perfect, but the larger tentacles can’t be inspected up close. Additionally, the tentacle texture can be quite low in resolution and still look great, as it doesn’t have sharp details that would fall short otherwise.

The tentacle textures were done inside Substance 3D Painter. First, I baked the textures using the high-poly Geometry Nodes version from Blender. My texturing process is pretty straightforward. In Substance 3D Designer, I made a very simple tentacle material with a few exposed parameters to use as a base inside Substance 3D Painter.

In Substance 3D Painter, I added a purple gradient going from bottom to top and hand-painted some yellowish blobs of color for some nice highlights. 

For the normal, I used some kind of detail normal from a premade Substance 3D material to give it the look of fiberglass, which is often used as a material for those attractions in amusement parks. You most likely saw it in a dinosaur park as a kid. 

B Bombsite Art Pass

Lizard: While tackling the art direction for the indoor areas of the castle, I wanted to make sure they also brought a lot of contrast compared to the outdoor areas. So, in opposition to very sunny and bright outdoor areas, the insides of the castle are very cold and raw. Everything is seemingly made out of stone and tiles. There is an occasional wood for the ceiling and support beams, but it's very limited. The lighting further amplifies this by having a cold temperature.

Production-wise, sadly, B bombsite was a part of the map that I was dreading to start to work on the most. I pushed it further and further down the timeline while adding details to other areas.

This is what the inside of the castle looked like on the 8th of February. The deadline for the next stage was the end 9th of February. So we technically had two days to finish it because of the time zone difference. 

It was a very stressful period of the project, but I would say the most fun at the same time. Every few hours, we would sit on the Discord server and brainstorm ideas.

Luckily, the structure of the bombsite was very easy to work with since the main room was just one instance of a wall rotated around. I had to art up that one instance, and everything propagated nicely.

It was the same case for the arches and big pillars. I only had to art up one arch and pillar and add some smaller details and material shifting to avoid repetition later on.

Working on such a big area so close to the deadline also forced us to do some “production optimisation”.

Radu: One of the ideas we had during the initial brainstorming for the project was to have a ceiling and windows inside the castle that would change their material either to daytime or nighttime based on a trigger that the player could interact with.

Although we didn’t quite deliver on this by the end, the idea steered us in the direction of having a rotating dome, set up by Lizard, with a night sky to contrast the daytime setting on the exterior of the castle.

The first dome material we used was done by Inez (Lizard’s fiancée).

We replaced the initial material with a version that I put together sometime at 5 am during the last day of the second stage. 

During the last days of the third stage, I improved the material and created a separate layer of clouds that rotate in the opposite direction compared to the dome rotation.

All of this was done by hand in Photoshop by painting the sky and the clouds on different layers while keeping in mind the rounded nature of the dome. After the colour bases were done, I used them to create the different masks for self-illumination and transparency.

Flo: Inside the castle are a bunch of corridors and rooms. On the side leading from B to A is the dungeon area. This area is supposed to be spookier with cages and chains hanging from the wall and blue and green lighting giving the place a mystical vibe. We tried to incorporate a lot of utility assets to further signalize this being a real place, which it lacked after stage two.

Worldbuilding

Radu: To support the theme of the map, I planned several golf courses through the gameplay and background areas.

The final addition to the courses came after the contest with functional golf courses that were implemented by Lizard and Flo.

During the final stage of the contest, to better sell the fantasy of the players going on an adventure, I wrote and put together a series of story panels next to each course in the gameplay space so that interested players can piece them together.

The basic version of it is that new players are heroes who answer the call to adventure and embark on a journey to find the holy grail in order to banish the evil giant octopus.

Other supporting elements of the theme included signs, which I was responsible for blocking out and working together with Flo to finalise them.

I designed a park map info panel that Flo turned into 2 distinct versions that we used to direct players to the 2 different bombsites in the map, a post with arrow signs that we reused for orientation within the map, a hanging cloth to highlight the start of a golf course, series of panels to show the name of the areas and the winners podium.

Final small additions included featuring a small shop with the golfing equipment in the middle of the map and setting up infrastructure signage like exits, toilets, direction arrows, etc., and safety props like fire extinguishers, smoke detectors, fire alarms, and so on, throughout the map to ground it as a real place.

Fun Facts

Flo: The creation of this map was a long journey with lots of headaches and funny bits in between. I wanna dedicate a small part of this article to some of those hiccups. 

The prop that took me the most iterations, time, and probably a few years of my life expectancy was actually the stairs kit that is used around the map. They always had a different problem, be it small holes between cracks, the wrong size to be modular, or they outright just looked bad. In the end, I accomplished a good final look, but it took a long way towards that, and I’ll never see stairs the same ever again.

Lizard: For the longest time, we didn’t have ideas and energy to figure out what the main hero prop inside the throne room should look like, so we pushed it down the timeline with the hope that we would figure it out at the end. Turns out the end came quicker than we anticipated, and we ended up with a fully arted bombsite and a hero prop that looked like this.

Literally a few hours before doing a final build of the map, I suggested that we should just cover it with some kind of tarp and call it a day, in hopes that we advance into the next stage, where we can finally replace it with a proper asset.

Flo exported the throne geometry to Blender and created the best-looking tarp I have ever seen while still maintaining the overall silhouette of the prop, since it was a very important gameplay piece as well. 

We decided to leave the blockout geometry underneath the cloth, and to this day, no one has noticed.

All of the swinging signs and cages on the map have physics enabled and can swing when shot. At the time, I thought this would be a cool feature. It turned out to be a huge time consumer and required a lot of debugging. I don’t think anyone even noticed it in the end, oh well. 

Closing Thoughts

Grail started by the sheer coincidence that Lizard and Radu had roughly the same idea for a contest map. Collaborating on it allowed them to throw a bunch of crazy ideas at each other and push one another to deliver something that had never been seen in Counter-Strike before. And once Flo joined the project, the vision for the map really started to materialize.

Throughout development, the team had to make bold and risky decisions while in doubt about how the map would be received. It was a huge group effort that paid off and created a strong bond between its creators.

The map won 1st place in the Big Adventures Mapping Contest and was added to the game as part of the official community map pool rotation on 8th of May, 2025. 

It was featured in the then-first-ever weekly mission of Counter-Strike 2. Over 1 million people played the map during the first week, and as of now, it has had almost 2 million players. It is steadily reaching around 700k players every month based on data from CSSTATS.GG. 

Looking at the statistics, Grail has managed to find an audience among the casual players of the game, which is what the authors intended from the beginning. 

As the Steam Workshop description reads, “The map itself was meant to be a callback to the creativity and unusual designs of Counter-Strike 1.6, when people explored wild ideas that pushed against the norm.”

The team is thankful to everyone who played Grail and the ones who uploaded videos, clips, screenshots, and posted articles about it.

It’s safe to say that Grail made an impression and proved that sometimes being bold can be a winning strategy.

Florian “Flo” Wagner: LinkedIn, Twitter, ArtStation

Radu Tanasie: LinkedIn, Twitter, Bluesky

Bartlomiej “Lizard” Guzek: LinkedIn, ArtStation, personal website

Interview conducted by Gloria Levine