Today we'll be talking about 'level design' - a core part of Project Argo. We hope you've enjoyed this week's live stream with us, and in this devblog, we hope to give you some more information about our work. Since level design is a rather large topic, we decided to write this blog as a team, with each section covering a different aspect of level development. First up, Jaroslav will talk about how to become a Level Designer, then Hedrik will explain more about what's involved with designing locations, and Jan will close off by explaining how we analyze player data and how we incorporate your feedback. Before we start, we should point out that all of us joined Bohemia Interactive this year, and that we've been working on Project Argo ever since.
There are essentially two routes to becoming a Level Designer: you can study at specialized school, or learn it on your own. While the first option is only possibly for some, the second one is available to everyone. First, you need to find a game that you like and which provides all of the necessary tools and at least some documentation, either written or in the form of video tutorials. Nothing can teach you as much as a tutorial from a modder who walked the same path as you one or two years ago.
The initial steps might be the most difficult. You have to learn how to work with the available software and explore the game mechanics. As a player, you probably already know how the game works, what the rules are, and what is possible to do. As a Level Designer, on the other hand, you have to create these rules, and decide what is possible and what is not. You learn as you progress - what seemed impossible on your first try, might become easy after some time. You will make mistakes, as everybody else before you, including myself. Don’t hesitate to re-do your work. Every hour you spend on designing levels will develop your abilities and of course your levels themselves.
It can take months, perhaps even years, before you finish something that you feel is worth publishing. You can release your map in beta, and continue to work on it - or simply as final. I recommend the second option, because it enables you to learn without any external pressure. That said, in my experience, most players out there appreciate modders and their work, and their feedback is almost always a positive thing. In that sense, it’s also highly recommended to play your 'level' with them. Many server admins are eager to host community-made maps, and since they're often veteran players or even modders themselves, they will give you some of the best feedback.
When you release your work, be prepared to fix the most obvious issues, and where possible react to player feedback. Once you feel that you have nothing more to give to your level, sum up what you think you did great, what went wrong, and then don’t hesitate to start designing a new level. After publishing a few levels, and hopefully receiving a lot of positive response, you might be convinced to try and apply for a job as Level Designer. I myself went through this, sometimes uneasy, path, having been a modder for years, before finally turning my hobby into a job at Bohemia Interactive.
Jaroslav Řízek (Level Designer)
In this next section of our devblog, I will take a rather broad approach to level design, and provide some examples using Project Argo. I will explain some possible techniques that I think apply to any kind of game, but also some of the specific problems that we had to overcome during the level development of our competitive multiplayer first-person shooter.
The creation of a level always starts out with an idea, your idea. You might have seen this really cool-looking building on a recent city trip, or perhaps you went to this stunning nature park once and thought “I would love to play in this”. Or you can look to various forms of fiction for inspiration; books and films can generate great ideas. And ideas are awesome! They can be the sparks that ignite an amazing fire, but you'll need to cherish them so they don't burn out.
Gathering sources can help you to get a hold of your idea. For Project Argo one of our best sources of reference are satellite images of the world. However, you can also take photos of nature and objects yourself. Or look for pictures online - basically anything that will give you some visual references. Once you've gathered all of this, you'll want to put it all together in a place you can access easily. I recommend making a moodboard, which would show the things that triggered your best ideas: that giant cliff on which you could see players fight on several height levels, or the harbor where you saw players battle over ships to capture valuable cargo!
Once you have a sense of what your level should look like visually, you will have to start thinking about how it will play. I strongly recommend you to prepare a plan on paper, before you start making something in-game. This will help you save time in the long run. A simple change on paper can translate into three times the amount of work when making that same change to in the game. So, first prepare 2D sketches, and if needed 3D sketches as well.
Will your level have a lot of height differences? Will it be open, or will it have 'tunnels'? These are the kind of questions you have to ask yourself when you start sketching. It's important to divide your level up into 'combat zones'. These are areas of variable size where players are forced to clash with each other. You want players to be able to move inside them and have a clear sense of direction when doing so. I like to use some rules to keep my areas consistent and fun to play in. Ideally they should have 3-4 entrances, but never less than 2 or more than 5. Players should see at least one exit from the point they entered to prevent them from getting lost. Create some cover with an 8 figure around the pieces so that players have opportunities to flank and move around. And last but not least, make sure that all your areas have their own unique elements befitting the level's theme. Going for a huge mountain? Perhaps have a crashed helicopter somewhere that players can fight around. And while you're at it, why not add a waterfall as well? These are all very recognizable features that players can call out to their teammates and helps them remember how to to navigate through your level.
You then want to connect these areas together. You can do so by simply adding hallways in between these places. But always try to challenge yourself. Wouldn't it be more fun if you had to jump down a low cliff to enter the next area? As I said before, it's easier to change something in the layout now than when you are halfway in building the prototype.
When you're happy with the sketches of your level, you can start prototyping. With Project Argo we were fortunate enough to have access to the vast library of Arma 3 assets, which helped us prototype levels very quickly, but also limited us in the way we could be creative with shapes. If you're using other technology and tools, such as Unity or Unreal, I recommend using simple cubes or brushes to create your prototype. Don't focus on the aesthetics just yet. The biggest hurdle to overcome here is scale. You might have designed the perfect solution in your sketch, but when you see your concept in 3D, you might find that it doesn't really work. Try to be creative in finding solutions, maybe add some objects, or remove some, and slightly adjust shapes and sizes. Along the way you might even find better solutions to the problems you encountered earlier. Include those in your first prototype, so that you can then start play-testing the overall level.
You've now reached the most important step. Play your new level, and identify what should be improved. For Project Argo, we take about an hour every day to test our game. By doing so, it becomes much easier for us as Level Designers to see what works and what doesn't. Aside from that, we also receive valuable feedback from colleagues on how to improve our levels. Are people having problems navigating your level? Then maybe you need to make sure that entrances, exits, and cover become more clear to players. Are players not able to find each other? Then your level might be too large. This is the time where you need to be critical towards yourself and focus on making improvements on all fronts. If your level is not fun to play now, it will probably also not be fun when it has all of the art assets in it. First make sure you are happy with the gameplay, and then can focus on making it look beautiful. Then, once your level plays well, and looks beautiful, you can release it to the public and see how the wider audience responds to it.
Hedrik Offenberg (Junior Designer)
After our levels have been made available to our community, we cannot wait to see how they will be welcomed and explored by players. To help us analyze the behavior of players, we collect and explore anonymous player data. The purpose of this is to learn from our mistakes and successes. The current pipeline for that is quite simple but effective. A quick overview of how it works can be seen on the image below.
To help you understand it all comes together, I've added four case studies to illustrate how we're currently improving the locations in Project Argo. By the way, all of these changes have made their way to Project Argo's Christmas update, which we released last Wednesday.
La Trinité is one of our largest Clash locations to date, and there's a lot of variety within its objectives. From mansions to vineyards, and from construction sites to an industrial area - all the way to even a football field; La Trinité has it all. Consequently, players enjoyed a mixture of open areas and more enclosed urbanized areas. This case study is about the Western base of one of the two fighting sides: Alpha.
On Image 2, you can see the base from a top-down perspective. On Image 3, we overlay it with the approach routes we wanted players to follow. We used all means at our disposal to guide players to the entry point of the objective.
On Image 4, you can see the player movement data that we collected from the week that La Trinité Clash was available. It's been filtered to only represent the rounds where Alpha was the objective. We were happy to notice that players were indeed following the approach routes (Image 5) that we wanted them to take. Given the rather open nature of the area surrounding the objective, guiding players was quite the challenge.
With Clash St Marie, we did run into some design issues. We knew that the rural, non-urban valley would present a bunch of challenges. Fortunately, some of these were fixed before the level was released to the pubic, thanks to our internal daily play-test sessions.
Foxtrot's first iteration turned out to be a troublesome maze. If you look at Image 6 below, you will notice that exits and/or entry points were placed illogically for players. Players sometimes had to run away from a certain objective before actually being able to actually start attacking the objective. It took players several games to get used to it, which resulted in players avoiding this objective completely.
So we went back to the drawing board, and we started discussing about how and where to fix this specific objective. We focused on creating a 'readable' location by limiting the amount of props, which made it easier for players to call out directions and/or the positions of enemies. The location also lacked memorable landmarks, so we then added a couple of very distinct props to the scene. The end result for the prototype candidate sketch is shown in Image 7.
The result is the version that made its way into the Open Prototype.
So, this first public version of St Marie went out the door, but received lots of mixed feedback. There were balancing issues, The north-western side was favored over the south-east. So we went back to the drawing board again. We'll showcase two out of the many resulting design improvements below.
At Alpha, defenders found it hard to figure out where attackers were coming from. Attackers also took very long routes, which can be seen on Images 9 and 10.
Improving the design meant cleaning up the forest and providing more guidance to both attackers and defenders.
We created open areas (see the Green "X" signs on Images 12 and 14) to try and demotivate players to cross these areas. On top of that, we added orchards and the results of this can be seen on the Images 11,12,13, and 14 (all of these improvements made it into the Open Prototype's Christmas update).
The data (Images 15 and 16) showed us that players approaching Charlie from the Eastern FOB objective (Foxtrot) had a more direct and covered route, which clearly benefited their side.
The design improvement was to provide the attackers coming from the construction site objective (Beta) with more clear, better covered, and diverse approach routes. The process can be seen on the Images 17,18,19, and 20 (these improvements also made it into the Open Prototype's Christmas update).
Jan Van Hassel (Level Designer)
As you can see from the case studies, Project Argo's terrain 'Malden 2035' is constantly changing. Even though there are some limitations in what we can do with this prototype, we're introducing improvements as we learn more and more about the kind of locations that Project Argo requires. Our relatively small Level Design team is constantly exploring new and interesting ways to enhance our locations and data-analysis pipeline.
Last but not least, we would like to point out that - even though we collect anonymous player data to improve the game - you as a player can also provide us with even more detailed feedback by posting videos, reporting issues to the Feedback Tracker, discussing your experiences on our forums, and via other (social) media channels. You have already been doing this so far, and it will continue to help us move forward with Project Argo.
All in all, we hope that we were able to shed some light on how we handle the development of Project's Argo locations. We're a young, highly-motivated, team, and we're working everyday on creating the best experience possible for you, the players. We thank you for the interest and support!