DOI Level Design Guide
September 15, 2016
Hello, I’m Jeroen Van Werkhoven the Lead Level Designer at New World, responsible for the output of the level design team. I work in close collaboration with our game director to maintain design vision. I relay these goals to the team and provide them with the resources they need to fulfill their milestones. In this blog I will give you a peek behind the curtain of the design process for the maps in Day of Infamy, and the choices we made to achieve our goals in the level design department. I’ve asked our team of level designers to also share some tips for those of you interested in level design or participating in our map contest.
Early in development, we decided we wanted to do something different for our Day of Infamy levels. We didn’t want to copy our Insurgency level design fundamentals and call it a day but try something more unique and fitting for the WWII setting.
To determine how we could accomplish this, we analyzed Day of Defeat Classic and tried to figure out what elements are necessary for our levels to give it a bit of a classic DoD feeling but still maintain our vision for the levels. While we were analyzing these maps, we noticed that they had a lot of adventurous elements. For us at first, that was a bit strange to implement in our game worlds, because for Insurgency we always try to go for a more realistic approach. For instance, in INS the player doesn’t have to jump through a window to enter a room often or walk across a small beam and jump to access a roof.
For DOI, we decided to add some of these described situations in our maps. Especially for our diverse class selection, it felt like the right approach to convey a bit of that adventurous gameplay into the DOI levels. We also noticed that DoD maps had a lot of alternative routes to flank an objective or just sneak past the opponents and pop-up behind them. And let’s not forget the accessible buildings, not every building is fully open in DoD Classic, but they created the feeling that the maps were less linear than they are. We tried to maintain this same feeling in our maps and pushed it a little bit further by making the majority of buildings accessible and on top of that we also opened a lot of rooftops to engage more vertical gameplay.
A player ability like artillery is a completely new design challenge for us. The level design team had to make sure artillery would work well on all our maps. One of the things we noticed with our traditional objective placement is that artillery could easily wipe out an entire team while trying to capture the point. To counter this, we made the objective zones larger and added additional protection such as a small building or roof segment, just enough to give the heroic soldiers a chance to succeed in their mission. When the sky clears, the player should have that feeling of “WOW.. did, I just survive that!?”
Players love running around with a flamethrower (including me) but as the flamethrower class, you don’t like to get shot all the time. We noticed early in testing that this class was easy to kill when they went for a frontal attack. Even though most players know that rushing into an enemy base with a big flammable tank on your back isn’t a good idea, we decided to add a few small adjustments that would make Mr. Burner a lot more fun to use. The key for this was to add a few sneaky spots and additional cover that could be utilized by the player as ambush points. This made for a warm welcome for the flamethrower class. We didn’t suddenly add cars and big props everywhere because that would break balance for other classes, but a few sandbags or crates somewhere can make a huge difference. As a tip, make sure to check your corners because there could be a.. “Aaah Aaah”….hiding.
Day of Infamy includes five game modes for Multiplayer and Cooperative:
When we were working on the DOI mod for Insurgency, we added our popular Push mode from INS. Even though it was working pretty well for our early DOI levels we decided that we wanted to do more with this game mode. We weren’t directly thinking of changing the rule set but more about what we could do from a level design perspective. So we went back to the drawing board and decided to add this mode in our current set of DOI levels. Adding Offensive mode was quite a challenge, because with a few exceptions, our INS maps are a lot larger in scale. The key for us was to treat it as an entirely new mode. First, we added the three objectives on the map. Placing the objectives and spawns was a bit of a puzzle because the objectives had to be a lot closer to each other due the compact level layouts. And the spawns had to be in a decent range of the capture points but without making it easy for spawn campers to sabotage a team. It took us some time play testing to figure this out. We didn’t want the objectives too distant or too close to each other, it must be balanced. Our goal was to have a constant level of action in this mode, to maintain the WWII feeling and make it more intense for both teams.
Once we were satisfied with the objective placement, we had one more job to do: adding the last stand. In the beginning, we didn’t have a destructible objective yet. But we soon arrived at the idea of adding the radio. However, it wouldn’t make sense that only explosives could destroy a communication device. So we decided to make it very fragile, just one shot from a pistol is enough to kill the radio. During testing, we had a lot of fun either defending or attempting to destroy the radio, but we forgot about one thing: Artillery! The radio’s worst enemy. A radio isn’t able to hide or run away, it just stands there there waiting for the inevitable end of its broadcast. The solution was to move radios inside a building, but make sure there is a ceiling on top to protect it from rain and artillery shells.
At first sight, Liberation looks like our leading three objective Firefight mode for INS. But under the hood, there are some significant differences to strengthen the WWII setting. The objective nearest to the spawn will give a team one wave when captured. But if the team takes B at the map’s center, they will receive 2 waves. The last point closest to the enemy base will provide the team with 3 waves when taken. Because of this new mechanic we had to uphold the same pattern in the layout of our levels, making each objective well balanced for both teams. Objective A and C should be close to team 1 and C close to team 2. Which means if a player wants to capture the point they have to go into the enemy territory to reach it. The objective should have enough flanking routes to take it, but at the same time the team that is spawning near the objective should have quick access to this area. Objective B is central in the map and should be perfectly balanced for both sides. Anyone who played INS will notice that our capture areas are generally a bit larger in DOI. We did this to counter the artillery and give a team a bit of a chance to capture the point even if an artillery shell hits directly on the target. For example, players can hide under the bridge in Comacchio when they hear that nerve-wracking sound of incoming shells in the distance.
Frontline is inspired by a game mode we had way back in the Insurgency mod called Battle. We determined this game mode would be great to implement for DOI and make it a bit more different and more fitting for the WWII setting. At first, it was a tricky game mode for the level design team to implement; it was essential to pick objective locations for B and D in the maps that had a lot of flanking routes. At the same time, the structure or area should feel like a front line which includes strong defensive positions. Objective C needs to be in the middle of the map because once either team has captured a point the spawn locations will change according to the faction that has seized the objective. When the advancing team takes the next objective, the radio will appear on the map and can be destroyed to win the round. But if the team defending their radio manage to hold off the advancing team and recapture the point, the front line will shift again. The doomed team always has a chance to make a comeback. The choke points on the map are always shifting, which made it a challenge for us to find suitable locations for the objectives B and D. As always, it took us many test sessions before we are satisfied with the result.
Our cooperative modes are unique in their own way. In Entrenchment, the players need to hold each objective as long as they can. Each time they lose an objective, the next objective is activated and the defending team will need to retreat. We noticed that it’s important for objective placement to select areas that have a good line of sight and sturdy defensive positions. The final stand requires a lot of flanking routes to overwhelm the defensive team.
Stronghold feels like playing a short campaign. Each time your team captures an objective the spawn location will advance further into the map, until the final radio objective that needs to be destroyed. We noticed for this game mode, the distance between objectives is crucial. If they’re too close to each other, it would be too easy for the defending team to protect both points. And of course, the radio needs to be well protected.
For Day of Infamy, we wanted to create both original maps and remakes of classic Day of Defeat maps. However, we determined that we didn’t want to create 1:1 remakes, but rather add our own vision while still reserving the gameplay elements of these classic playgrounds. It put a lot of pressure on our level design team because we knew doing a Caen remake is not something to be treated lightly. Many old school players remember those epic battles like it was yesterday. But at the same time, we also wanted to add something new for our younger players and let’s not forget our INS player base. So we had to find a middle ground.
At the beginning of Ortona’s design (known then as Catania) we weren’t entirely sure how to achieve the middle ground, and just started with a block-out faithful to the original Caen layout. It took us several weeks to have something playable for internal testing. During these tests we noticed early on a big flaw: the map was too big and players were scattered all over the playground. Also, our symmetrical game modes weren’t much fun on this map due to its linear layout. We decided to do some drastic changes and cut the map by about 40%. We also adjusted the overall design and made it more compatible with our other modes.
We were very pleased with the results of our playtesting and the iterations made the map more balanced and fun to play. However, because of the significant tweaks we did, the map was lacking some crucial elements that gave it that iconic Caen feeling. Once again we analyzed the original map and added back vital parts to ensure Ortona would feel more like Caen. It took us many test sessions before we were happy with the result of the layout and at that time we had only a few weeks to complete the map for early access.
Another map we created a remake of was Ramelle (later renamed to Merderete). After the first block-out, we weren’t very happy with the result; it felt too much like a copy of the original map. So we went back and instead of just changing it, we searched online for Italian places with bridges. We decided that Comacchio was the perfect fit for our Ramelle remake, it has a bridge and multiple church towers which are the necessary parts to give it that classic Ramelle feeling. We maintained most of the original layout but added more additional paths, and as with Ortona, moved things around a bit. One of the major differences is that we added more vertical gameplay to the map by opening rooftops. Our testers were very fond of this change. Another famous part of the map is the bridge and river. Already early in testing, the river had an enormous impact on gameplay. Most players tried to avoid it because once they entered the water, they were pretty much sitting ducks. Although we weren’t against this feeling, we decided to make a few small but crucial adjustments. We lowered the river a tad so players could more easily move through it. We also added passages in the form of stationary floating boats that could be used to cross the river on both sides near the team spawns. It took us many iterations before we were happy with the design of these so called bridges.
Maps from DOI mod
For the Day of Infamy mod, we already created Bastogne and Dog Red, but it was too easy for us just to polish them up and release both maps in Early Access. Instead, we analyzed both maps and made some drastic changes.
In Bastogne, we changed the lighting to make it less white and more atmospheric. Our primary gameplay concern was that the foxhole objective in the middle area was too deep in the ground. It didn’t feel like an objective and it was very hard to defend. This led to the decision to overhaul the entire center area and move it to higher ground. Another big part we changed in the map was the town. For the mod, it was pretty much a bunch of grouped buildings, didn’t feel much like a town and also the gameplay wasn’t optimal. So we removed nearly all the buildings except the church and redesigned the entire street plan to make it not only feel better as a town but also to make it a lot more fun to play in.
Dog Red was already a solid map in the mod. However, we always had the feeling that A in the Offensive mode wasn’t in the best spot. The area didn’t feel like an objective, so instead we moved it to the beach bunker. At first we weren’t sure about this location during our test sessions because it was almost impossible for the Allied forces to capture it. But just reverting it wasn’t an option. Instead, we added another entrance at the foot of the beach bunker which improved the entire level flow around that area. Next, we drastically adjusted the back area of the map. We added more cover initially, and later on after we released early access, we also extended the map and added an underground tunnel system. This made a huge difference for the overall level flow but not directly in a positive way. We noticed that after A was captured everyone was pretty much in the tunnels. So to avoid this from happening we cut a large number of them and moved the tunnel exits to more open areas. The effects of this change are noticeable so far, and players have to expose themselves now once they get closer to the mainland.
For us, it was important to maintain a consistent quality level throughout our maps and give each the care it needed. But at the same time we felt that our game could use some additional content. We know players like a wide variety of different maps they can play. However, designing maps is a lot of work and in particular, the block-out stage is very time-consuming and challenging. For this reason, we decided to use Verticality and Sinjar from Insurgency as a base layout for our DOI maps Sicily and Reichswald respectively.
The first new level we wanted to create was Sicily. We already had a good idea for the direction, but we had to find the perfect candidate in our level collection. It didn’t take long until we decided to use Verticality as a base for Sicily. Even though it was far from perfect, we were able to use quite a few buildings and a large part of the terrain. Using this workflow helped us a lot to test Sicily and later Reichswald in the early stages of development and also gave us some direction for the visuals. Though after we had established gameplay, we overhauled every part of the map to make it consistent with our Italian level theme. We didn’t leave a brush untouched.
Our forest environment for DOI, was a bit of side project for us. We weren’t sure at first if we could make this map happen in time but we had a good idea about the theme and gameplay. We just had to find the right fit which happened to be Sinjar. It took us by surprise too but we wanted to create a large-scale map, and Sinjar is our biggest map in INS. So, after all, it wasn’t that strange to us to use Sinjar as a base for Reichswald. Also, the big hill was perfect for the first Offensive objective. Even though we copied the terrain from Sinjar, we still did a significant amount of displacement work to make it feel more like a European landscape instead of the rugged middle east.
Design by Data & Feedback
Creating a well-balanced level is a huge challenge. You can create the best-looking map ever but if the gameplay isn’t fun players will drop it soon after release and never come back. So for us, it was fundamental to put all our focus on the gameplay first and try to create a fun experience for our players. It took us a lot of testing and tweaking until we were satisfied with the maps for a first release. We had six maps in our early access launch, and the first impressions were great, we had a lot of positive reviews on Steam. So you would say, time for a holiday!
Well not really, during the first week of release it’s especially important to keep a close eye on the balance of levels and any bugs that might pop up. While of course we tested our game an enormous amount of hours internally, the playtime a map receives on the first day of release is something we can never surpass with our testing team. It’s very exciting when a map goes live for us level designers; it’s always a bit of a gamble if certain gameplay elements in a level work or not. As a creator, you always have to trust your instincts and use your experience to hopefully create a map that is well-balanced and most importantly pleasant to play. It’s important when a map is not performing well to determine what the issue is and how to improve it without tearing up other parts of the level. Because a small adjustment in an area could break the entire level flow in the worst possible case. We always take a lot of time to analyze gameplay on the levels and to make a plan for improvements. We’re not always adjusting everything at once because we need to ensure what we improve is an improvement and not just different. It’s something a level designer always has to ask him or herself when iterating on a level.
We’re very grateful for the feedback we have received from our player base so far. It’s extremely helpful to us to improve our levels and make them the best as possible. We’re planning to update our levels through early access and beyond if necessary. And with the introduction of our “In Testing” playlist we’re able to have our players test run new content at an early stage. We are also now relying upon the survey for anecdotal data and in-game statistics recording to determine balance. Exciting times ahead!
To make sure we were able to finish all the levels in time with our small team, we decided to introduce a “carousel” system among our level designers. Every level designer would work on several maps during development. For us, this was the right approach to pump out the needed content for early access, even though in the Source Engine it’s always tricky to work with multiple level designers on a single map. But by communicating often, we were able to make a lot of progress in a short amount of time. Fun fact, not only our standard level design team worked on the maps but also our CEO Jeremy Blum and producer Jeremy Faucomprez have been hammering. As you can read, creating DOI levels at our studio is a team sport, and we had a lot of fun together developing these maps for you.
Our level design team is comprised of different levels of experience and skill-sets, but all with one goal: to create high quality maps for our players for the time to come. I asked our level design team to write a few paragraphs about the work they’re most proud of, and give some advice for beginner and experienced level designers out there. We’ve only scratched the surface of possible WWII theaters, and we are more than eager to see what you will create!
Steve “Chief-C” Cron
“Hello I’m Steve Cron also known as Chief-C and I’ve made numerous maps for Insurgency, but Day of Infamy was my first venture as a professional. I learned early that I had been making some fundamental mistakes when working with the Source engine. If you’re thinking about making a DoI map, keep these developmental standards in mind.
First, don’t polish the map as you create it. Always create a “block-out” version (very little art and few props) and test it. Getting a map properly balanced will require numerous layout changes. Using a basic block-out map will save a great deal of work. Place windows and doors, but keep it very simple. Only use the props necessary for game play (cover/concealment).
Always keep your brushes snapped to grid. In addition to that, don’t set your grid to the 1 scale. Unless you’re placing props, never set your grid to the 1 scale. Did I mention you should never set your grid to the 1 scale? Don’t do use the 1 scale… it kills baby kittens.
Use a standard height for walls, floors, doors, and windows. The props will dictate window and door sizes, but also consider the height off the ground. The perfect cover height is 48 units and allows bi-pods to deploy easily. For DoI, our standard bottom floor is 16 units, with a 160 unit wall on top of that. The second floor is another 16 unit floor, with a 144 unit wall on top. The props we’re creating will work well with those dimensions, so I recommend using them. In addition, always make your walls at least 16 units deep. Players will appreciate it and pay reverence to your mad map making skills.
When I first started making maps, I had no clue what size to make my buildings. I figured I would just duplicate actual building dimensions and it would be good. Unfortunately, most buildings aren’t designed to have 32 super-rugged manly men fighting through their hallways. Don’t make your environments too tight. Also, try to have at least three independent routes to the objective. While I enjoy a big pile of bloody and smoking corpses as the next guy, chokepoints don’t make for good game play.
Lastly, always use the Hammer “Check for Problems” function before running the map. Nothing makes me sadder than wasting 3 hours compiling a map to discover a few overlays weren’t assigned. Have fun and we hope to see some great custom maps!”
Zach “cincinnati” Snyder
“Hello, I’m Zach Snyder, a designer and artist on the team. My work on Day of Infamy has included the design of the cathedral on Ortona. I began work on this after the basic footprint and interior volume of the cathedral had been grey-boxed, playtested, and iterated upon by Jeroen. The challenge was to create believable spaces in an area that was much smaller than the typical European cathedral, and to convincingly allude to classical form while minding the performance concerns that too much detail presents. Most important, though, was the imperative that the spaces should encourage great gameplay.
Cathedrals are inherently modular buildings, so after finding repeatable dimensions that would work well inside the given footprint, we were well prepared to develop a minimal set of modular props that could be snapped together throughout the building. After initial focus on geometry that directly affects gameplay, and making sure we had what we wanted, some of the more cosmetic exterior features were added.
One tip I’d offer up specific to this junction of level design and environment art is to exploit the building type you’re using for what is important to the gameplay. For instance, it’s typical for the chancel/altar of a cathedral to be raised. This is exploited on Ortona to create a visual, physical, and psychological barrier between the two teams in what is essentially often the center of gameplay in several of our game modes. The horizontal plane of this platform corresponds with player eye-height on both sides, and an altar in the middle raises that height a bit more in the middle. The platform, in turn, becomes highly sensitive in regards to capturing the cathedral and otherwise moving through it.
Another exploitable aspect of the cathedral type is its central nave and side aisles. Rows of columns are one of my favorite gameplay devices, as they allow, via movement around them, nuanced control of simultaneous lines of sight. Depending on your visual angle through and physical proximity to them, they can range from being fully opaque walls to thin slivers of cover, making for some fun (and often funny) cat and mouse firefights. While a bit more substantial than columns, the nave and side aisles of Ortona’s cathedral cut three bands of space down the length of the cathedral, which are diced up by perpendicular bands implied by the arches that connect them, creating a dynamic grid in which you are never fully in cover and must stay on your toes.”
Brian “LATTEH” Birnbaum
“Hello my name is Brian Birnbaum and I am one of the level designers for Day of Infamy. I have worked on several of the levels and helped make most of the sky box textures for the game. It has been amazing working with the team on DOI, what is great is getting to create amazing things and working with everyone to make those even better. Making the sky textures has to be my favorite part of the entire process.
When creating the textures, I would start out with a reference image (either from one I found or was suggested to recreate). Then I would spend a lot of time tweaking the sky’s in e-on software’s Vue. After rendering out the images I then had to tweak the contrast of the images and edit them so the fog ingame will blend into the skybox. Overall it has been amazing making those skyboxes and giving the levels a very vibrant sky.
Starting out in level design is pretty difficult, there is a lot to learn and a lot to do. The one thing that has been invaluable to me and is very easy and is sometimes overlooked too. Play other levels! Either for the game you are working on or another in the same genre. The worse the better, find what they did right and wrong. Learning by the mistakes and success of others will help you possibly save some time from making those same mistakes.”