GRB - Creating Worldbuilding Content
What follows is a breakdown of my process when working on Level Design, seen from a Worldbuilding perspective, as well as two case studies of locations I worked on that took contrasting paths to their completion.
Process
Speaking generally, creating level design content for games follows the following steps:
-
High Level Context (Where)
-
Thematic Brief / Pitch
-
Concept
-
High Level Layout Lock
-
Layout Lock
-
Full Content Lock
-
Ship
High Level Context
Before getting started on our level, we need to have some context. How much is already known about the game itself: 3Cs (Character, Camera & Controls), story, setting etc. Where do we fit into the game, what part of the game are we making, is it part x in a sequence, or zone y in a larger whole, etc.
Thematic Brief / Pitch
Usually the core team will send out briefs that communicate their vision for what your content will be based on, so that each section can be differentiated and mesh well with the overall structure.
Sometimes you get a fairly blank slate and are invited to pitch your ideas instead, taking care to work with what’s known (biome of the zone, relevant story beat, gameplay goals, etc.
Either way, the Thematic Briefs are fairly broad, and it’s up to you as a Level Designer to interpret them in a way that will complement the designs you will develop. Use them to inspire your Concepts, but don’t let yourself be too restrained by the Brief if it hurts your Concepts.
Be brave enough to suggest things that are maybe adjacent to what was in the Brief if you think it can be a big win for the Level Design, just be sure to bring it up early in reviews and make your case with arguments that are supported by the projects design intentions.
Concept
The concept is the first crucial moment where your design chops are put to the test. Sometimes this step can be done with whatever means come naturally to you, be they pen and paper, Photoshop, Microsoft Office, etc.
Often times there is a template of some kind that dictates in what formats you can create your concepts. Personally I prefer the granularity of Photoshop and similar software to the more rudimentary approaches.
I’ve found that final designs remain heavily rooted in their concepts, which is why I take things like Metrics, Lines of Sight, Cover, Flow, Thematics and Landmarks into account when working on my concepts.
While there such a thing as trying to go overboard before the time is right, I believe it’s important to have an awareness of these design ideas while working on your initial concepts, to ensure you will not have to redo all your work at a later date.
When I work on a Concept, I always try to come up with a unique trait. In my opinion it’s important to always try to do something new when you work, so you can run into new problems and grow as a professional while figuring them out.
However, because the unknown is the great enemy of production schedules, it’s important to be able to scale your ambition to the time available.
It’s a skill that develops over time, but checking in with the core team early can greatly improve the odds of finding that balance.
High Level Layout Lock
Once the concepts have been approved after successive rounds of reviews, we start working in 3D to get our first sense of how accurate our concepts were. In my experience, the more verticality is involved in your design, the harder it is to nail your design in the concept phase.
Similarly, the less control you have over the props that will go into your level, the more vulnerable your level is to future metric issues. So it’s important to stress test both of these issues at this stage.
Get some example props in your early 3D block out to get a sense of how your metrics will play out. Try to incorporate your terrain or height differential as soon as possible to see how it looks in 3D.
The biggest stumbling block at this stage in my experience has been to mistake the metrics you see at these early stages for the metrics you’ll be working with once the game is nearly done.
If a room looks right at this stage, it’s probably too small, if a space looks too big, take some placeholder props and give it another look. To make sure all of these issues are taken care of, my order of prioritization looks like this:
-
Unique Selling Point
-
Thematics
-
Flow
-
Metrics
-
Lines Of Sight
-
Cover
-
Landmarks
The reason for this is quite pragmatic, each successive element has less of an impact on what’s above, meaning that it’s often easier to change the latter as opposed to the former due to what they influence.
This is not a one-size-fits-all prioritization, but it’s a fairly accurate oversimplification.
Layout Lock
While every dev cycle is different, there is a good chance that most of the gameplay objects will only start to coalesce by the time most of the layout has already had to have been locked down.
This means that you will have to move / delete props and use the open spaces you have left in order to place your gameplay objects.
If you planned ahead correctly, you were able to get some placeholder metrics for these gameplay objects in order to integrate them smoothly into their surroundings. Failing that, you might be able to make some fairly accurate predictions for the gameplay needs, and leave some empty spaces for them in your layout by the time they are available.
I’m the kind of Designer that likes to have a high visibility on what’s happening in the project, so that I can anticipate these kinds of issues, but also so that I can flag important unknowns that haven’t been figured out that have the potential to impact our work.
To that end I try to spend my leftover time reading up on our next production steps, or if unavailable, research them by checking out what the teams preparing them are working on.
Other great uses of downtime are investigating shortcuts and bug fixes that can speed up the workflow of the team, or trading feedback with the rest of the LD team to improve everyone’s work.
Full Content Lock
After the Layout has been frozen in place, the attention of the LD has to shift to the gameplay objects and assisting the Level Artists, who will now have to shepherd the level to completion with your assistance.
Throughout you collaboration with Level Art, this is mostly about preserving the Design intent as much as possible, while responding to any issues the art team may run into.
Sometimes this can mean that LA uncovers that one of their intentions will have to impact LD in a big way to be accomplished, in which case it is up to the LD to find a suitable compromise that either preserves the design, or gives the design a new twist that allows LA to achieve their intention.
It’s all about communication and flexibility at this stage, so that everyone can achieve their intentions.
This communication and flexibility are also very much needed when working on the gameplay objects in your level, as there will be many points where other teams tweak their intentions, causing your map to require to be altered in order for it to suit the new direction.
The best way to safeguard against this is to maximize both your own visibility, as well as those of the other teams.
Let other teams know how their features are currently being used in content, and ask them for their input to speed up their iterations on their design if this is not yet part of their process.
Ship
The last stretch of development is an interesting time for Level Design, sometimes it means moving on to another project because there is not enough debug work to go around.
Sometimes it means holding down the fort with some debugging efforts while other members of the team move on.
If you happen to be taking care of content at this stage, it’s all about communication. Getting the right issue brought to the attention of the right people as fast as possible in order to keep the game as stable as it can be.
Resist the temptation to be frustrated when issues keep getting wrongly assigned, and instead focus on contacting the relevant team members to smooth out these inefficiencies in order to speed up everybody’s work, the end product will be all the better for it.
Server Farm Case Study
Let’s take a look at the Server Farm, a location of mine that was relatively complex to design, one which developed my ability to navigate various needs stemming from several different visions:
-
High Level Context
-
Thematic Brief / Pitch
-
Concept
-
High Level Layout Lock
-
Layout Lock
-
Full Content Lock
-
Ship
High Level Context
The Server Farm was located in one of the Highland Provinces, eventually called Silent Mountain. The province was pitched as the center of Information Technology, everything from Servers to Cooling Systems and the Control Centers to watch over it all.
As It was located in the center of the main island, it was to be a medium difficulty province without a large military presence, with only select points of interest that were to be heavily guarded.
Thematic Brief
The location itself was to be a Server Farm which centralized a lot of sensitive data for Skell Tech.
In terms of placement with regards to budget restrictions, the sheer size of this location meant that it dictated where a lot of other locations would be placed.
It’s importance also further motivated us to prioritize placing it in a location that was easily accessible and made sense thematically.
The concepts that were drawn for it had a minimalistic polygonal look on the outside, contrasted with a cathedral esque appearance on the inside.
The religious element was to tie into the theme of man revering the machine.
My take on the Brief was that I wanted to lean into the Polygonal look of the main building, and extend that design to the rest of the location, something which required a bit of coordination between myself and the Level Artist I worked with.
Concept
LDs were tasked with the prototyping of all the specific buildings that would only appear in a single location, of which this Server Farm was one.
During Prototyping, it quickly became apparent that the Aisles and the Nave of this “church” would be taking up the majority of the space available for the building, meaning that any other rooms planned in the building needed to be kept to a minimum and of a small size.
Once the Main building was handed off to the design team in charge of interiors, we started experimenting with the Polygonal style for the rest of the settlement to see what kind of outdoor gameplay spaces we’d be left with.
It started a trend for my design that would prove quite useful going forward, which was to design these locations primarily using the Negative Space left by the placement of buildings in the location.
The reason for this was because LDs did not have much control over indoor layouts, especially when Flow was concerned.
This meant that the bulk of the design work was to built on top of the pre-constructed building layouts, and focus the flow of the location on outside portions.
In this particular location though, the polygonal aesthetic meant that the inaccessible buildings we created were effectively created to fit exactly into the space left over from where we wanted all of our paths to flow into each other, which took a lot of experimentation and iteration to get right.
High Level Layout Lock
It was soon after we finished the Concept phase that we were informed that a Main Mission would be taking place in the location, and that there were some requests to be incorporated to accommodate it.
The request meant that I had to incorporate a couple of buildings into the settlement that would not fit in well with the Polygonal aesthetic we were planning, but these same buildings also needed to be partially hidden from view, this meant I saw an opportunity to kill two birds with one stone.
My Level Artist and I quickly began prototyping ways to hide this couple of Mission Specific buildings from view.
We tried creating a natural alcove from the surrounding rocks on the same height level, but soon found a solution that proved far more fitting: creating a man-made alcove at a height level lower than the rest of the location, overlooking a nearby lake. This meant we could still incorporate the roof of the man-made alcove into our polygonal style, while incorporating the new buildings into a smaller secretive section that was clearly separated from the rest of the location.
Most of the iteration for this location centered around the two features I’ve discussed so far, the Polygonal aesthetic and the underground section needed for a Main Mission.
Throughout the various reviews and rounds of feedback, various requests were made, particularly surrounding the underground section.
The flow into and out of the underground section was altered many times, as well as the style of the roof, going from a plaza to clear glass, and from a black opaque surface to a Polygonal Roof with a moderately opaque glass section.
There were many points where I was asked to either alter the underground section, or remove it entirely.
This really tested my willpower, as I was aware of the problems that led us to create the underground section.
Luckily, with some patient arguments and the flexibility to investigate all alternatives presented in reviews, we were able to show time after time how we could come up with a compromise solution that would solve the issues present with the current iteration of the underground section, without having to remove it or alter its purpose beyond what it was designed for.
It taught me to stick to my guns in a way that did not come from stubbornness or aversion to change, but instead from a level of awareness that surpassed those that were reviewing my work.
Specifically an awareness of the various impacts of potential solutions, impacts that helped convince my colleagues when labelled and explained correctly.
Layout Lock
Approaching the layout lock, a lot of our iteration was focussed on the high level flow of the location, meaning the entry paths into the location, as well as how they connect to the more central paths.
Because we created some additional inaccessible buildings in the polygonal style of the server farm, there was a lot of playable space that essentially turned into long corridors without many flanking options.
We’ve tackled a couple of these where it was possible by adding some additional smaller areas, such as the cooling facilities, the storage bay and the small roundabout. Another big improvement was the creation of the parking lot in the underground space between the server farm and the mission specific buildings, which also helped connect them from a logic perspective to each other and the corridors above.
The Parking lot was mostly created in order to facilitate an approach from what was deemed the most common approach angle, the south side.
We later also placed some rocks and a mountain path to further suggest the player with some options for their stealthy approach.
During this phase of development there was not a lot of conflict in feedback, it was mostly just flaws being pointed out and addressed accordingly as they came along. The biggest aspect of iteration on the gameplay side came from the AI archetypes and more specifically their amount and distribution.
The idea to have some snipers present was discussed quite early and remained the most stable element of the AI archetypes.
There was an idea to have a large amount of shotgun enemies because there was a desire to dissuade people from fighting in the underground / indoor areas as those were the most heavily guarded.
Instead we wanted to give the player an advantage when engaging in the corridors above ground.
The USP of the location was called “Kick the Hornet’s nest” to signify this idea of a strong indoor force pouring out of the buildings when engaged aggressively.
This idea was eventually abandoned when we learned just how oppressive the shotguns turned out to be, as well as a need to include a bigger variety of Archetypes, such as the Minigun Enemy, Drones and a Turret.
Because of this need for other archetypes, we decided to use them to reinforce the main angle of approach and persuade players to either sneak into the parking lot, sneak through the mountain path towards the cooling facilities, or make them choose another entry point.
Full Content Lock
There were not a lot of changes made to the overall flow of the location after Layout Lock, but there were two points of note.
We added props to make a connection above ground between the Mission Buildings and the Server Farm more pronounced.
Additionally there was a gameplay request to add breakable fences to our locations, which was incorporated in a way that opened up some additional entry ways into the settlement near the cooling facilities.
There were a lot of tweaks made to prop placement in order to create a more well-defined environmental story, as well as to excuse away the placement of various necessary assets such as the Civilians.
In order to provide players with more approach opportunities, as well as in order to props the Parking Lot, we also placed a Generator.
On the flipside, we also placed some searchlights as well as two AI archetypes that exist to make the players life a little bit harder, the Comm Enemy and the Retreater. The first of which calls for reinforcements if not taken down in time, the second of which attempts to flee the settlement once the player is detected, in order to remove a VHC from the equation and potentially alert nearby NMEs.
Ship & Conclusion
Nearing the end of the project, changes became more and more minor.
There were some tweaks here and there to spawning and props, but most of the work was centered around fixing the various issues that would pop up for one reason or another, in order to ship as stable of a location as was possible.
Looking back on the Server Farm, I’m really happy that we managed to hold on to it’s unique traits without having to compromise them beyond recognition.
While it’s not the most flexible location due to its lack of flanking options in the corridors, I feel like that risk was mitigated to an acceptable degree while allowing for a different flavor.
The Unique look of the settlement initially caused a lot of headache for my LA and me, but once we pushed through it became one of the pillars of the location. The same can be said for the underground section.
Having a Mission requirement really helped us to hold on to our ideas by giving us more leeway with Leads in discussions.
Working on the location helped me learn to distinguish between stubbornness / fear of change and a reluctance to let go of a solid idea.
It also helped my communication skills by learning how to vocalize that separation and articulate why the solid idea would be worth the additional trouble.
This communication was essential in order to convince various groups of people at various stages of development, not just for preserving ideas, but also for improving feedback received.
I was able to connect with a lot of validators, explaining my intentions in a way that allowed them to not just comment on what was present, but collaborate with me on how to achieve my goals.
The biggest downside of the location is an issue that has plagued most of the locations in the game: a disconnect between LD intentions for AI behaviour, and the final results.
Sadly we did not have a good line of communication with the AI team, and as a result, were mostly left to try and balance our locations despite the ever changing behaviour, as opposed to balancing the locations with the behaviour in mind.
Pumping Station Case Study
Let’s take a look at the Cold Water Pumping Station, a location of mine that underwent some radical changes throughout development, showcasing some lessons learned and the dedication to persevere despite setbacks:
-
High Level Context
-
Thematic Brief / Pitch
-
Concept
-
High Level Layout Lock
-
Layout Lock
-
Full Content Lock
-
Ship
High Level Context
The Pumping Station was located in one of the Highland Provinces, eventually called Silent Mountain.
The province was pitched as the center of Information Technology, everything from Servers to Cooling Systems and the Control Centers to watch over it all.
As It was located in the center of the main island, it was to be a medium difficulty province without a large military presence, with only select points of interest that were to be heavily guarded.
Thematic Brief
The location itself was to be a Cold Water Pumping Station, a facility in charge of converting nearby glacier ice into cooling liquid for the Server Farm located lower on the Mountain.
The location was initially cancelled at a very early stage, in order to correct our workload into a more manageable amount.
It was however reintroduced late in the Concept phase in order to reinforce the thematic of the Server Farm, as well as to make use of some leftover budget that was discovered to be unused in that vicinity.
My take on the Brief was that I wanted to make use of the nearby glacier, and attempt to incorporate that into our gameplay, which is something we would continue to experiment with for a long time.
Concept
Since this location was to be small and isolated, there were not a lot of resources planned for it.
No nearby infrastructure, no specific buildings, a low amount of AIs.
We planned to lean into that by creating a location that felt more like a collection of satellite locations, connected through natural paths over and around the nearby glacier.
We constructed a small outpost that fed from the glacier itself, and tried to terraform the nearby terrain into natural looking environments that could serve the rest of our gameplay needs.
High Level Layout Lock
Our initial approach all centered around this idea of a playable area that incorporated the nearby glacier.
We had some exposed parts of the cooling pipeline, some passage ways over the ice. All in service of this idea of blending nature with smaller man made satellites. Eventually we added a dedicated building that would serve as our central pump, which would house our sniper nest.
We started tweaking a lot of the terrain and rock placements in order to see if we could create a good flow with all this natural terrain.
The main struggle at this time was to make the location feel coherent, as well as to craft terrain and rock placement in such a way as to create cover without losing the realism appeal that the Ghost Recon outdoor environments have always prided themselves on.
This took up a lot of our time, and was just never really clicking in a satisfying way, although serviceable.
Layout Lock
Near the end of Layout Lock, it started to dawn on me that I needed to make a drastic change if this location was ever going to come together.
At the time, I was still hopeful I could create a nature focused location with the small amount of AI we had available.
I attributed the issue primarily to the unwieldy flow we had, as well as lack of well integrated Line Of Sight breakers / covers.
I soon found myself with a slow week on the LD side, and discussed with my LA to do some terraforming to explore potential major reworks for the location. It took me a day or two to get the hang of the terraforming tools, but I eventually created a good approximation of my desired layout with some help from my LA.
I spent the rest of the week bringing this next iteration of the location back up to par with all my other locations, placing buildings, props, AI etc.
I kept working on it for another week, tweaking it as much as I could, into the next phase of development.
Full Content Lock
Early on in the lead up to Full Content Lock, I was confronted with reality.
Sadly, and this is where I have to give credit where credit is due, it wasn’t enough, and my Level Design Director saw it quite quickly, while I was still struggling to hold on to hope.
There simply wasn’t enough Enemy Presence to warrant the size of the settlement, and the natural terrain meant that most of the paths felt like really long corridors without flanking opportunities.
While the second iteration had fixed the flow issue, and presented plenty of cover opportunities, it felt empty and uninteresting to play.
My director appealed to me from an artists perspective, noting how the many changes we’d made were impacting the perception of realism for the terrain.
It all became a bit too convenient.
He suggested to “repair” the terrain and try to replicate the location afterwards.
My LA and I took him up on his suggestion and decided to redo the settlement from scratch with this opportunity we’d been presented.
In order to make the most of it, I decided to pick a new Unique Selling Point, in order to grab the best aspect of the iterations we’d created up until now.
We went from the “Sneaking and Sniping” USP, to “Fortified Underpass”.
My LD Director and I both agreed that we should try to incorporate our most unique element, the Pumping Station building that was used as a Sniper Nest before, but reinterpret it’s use to be more central.
For me that meant building the new iteration around this building and all of its unique properties. I decided to create 3 layers of verticality that would integrate with the building, in order to make the most of all it had to offer.
My LA and I decided to drop the Nature and Satellites approach in favor of a more traditional man made location.
We chose to give the glacier more room to breathe by drastically reducing the playable space and leaving the rest of the glacier intact with a bridge connecting what used to be the main approaches of the location.
From a visual standpoint, we reused all of the buildings and most of the props, but placing each one in a new location that served the tweaked functionality of the location.
We traded the catwalks that were connecting the Glacier Tanks, and instead used them to create a more structured catwalk system between our Pumping Station and the Water Tanks.
In order to focus the experience of the player, I decided to enclose the location with fortifications that would motivate players to approach either from the main entrance or the helipad.
The result was a more consistent gameplay experience that ensured most players would enter the base instead of fight every opponent from outside the location.
Ship & Conclusion
During the lead-up to launch, most of the bugfixing on this location was centered around prop placement, usage and smoothing out any inconvenient metrics.
We worked on integrating it better with the surrounding road network, as well as some tweaks to fortifications surrounding the locations, but other than that it was quite stable.
After we made the tough choice to redo the location, not once but twice, I have to say it was for the best.
Spending the extra hours learning more about terraforming, experimenting and finally redesigning the location with the lessons learned, honestly created what I now consider one of my favorite pieces of work.
And while talking about it inevitably means admitting the mistakes I’ve made along the way, I feel like the lessons I’ve been able to learn as a result have more than made up for it.
While I wouldn’t say that what I attempted was not doable, I would say that the combination of a large playspace, a lack of man made infrastructure as well as a small amount of human presence combined to form a cocktail that just wouldn’t mix.
Having the guts to press the delete button on all of the work done on it for nearly a year is something I’m quite proud of, even if we had help along the way from an LD Director I respect a lot.
In addition, my dedication to unique traits in all of my locations is what in my opinion managed to elevate what could have easily been a generic hastily created location, into a unique experience that exposes the player to a new set of gameplay options and navigational opportunities.