Doesn't they say the same thing about photography when it was emerging? ;)
Agreed. This is just depressing and is a detriment to society. If this keeps advancing at its current rate, good art will be so trivial to generate that it won't be special anymore. Art will slowly morph into a banal distraction, with creating an original piece being as easy as applying an Instagram filter. The role of the human artist will change from a craftsperson to someone who picks a bunch of parameters, gives it to the AI, and chooses the best output. This type of technology is a threat to the very existence of art as a craft, will completely devalue artwork, and will make the journey of training to become an artist obsolete. I hate these researchers for what they're doing to a field that I love.
I disagree. There will always be demand for real artists. Like any other digital software, this is just a tool with the possibility to help artists create compelling worlds faster and add realism that would otherwise have taken days to make using other methods. As a 3D character artist, I would love to use this to create quick backdrops to place my characters in to enhance final renders.
Nathan Cheever detailed his approach to the creation and planning of video game worlds.
My art talents followed me through high school launching me into a Fine Arts degree. The impulse to explore other forms of expression led me to explore degrees in Graphic Design and Film. Then as computer graphics became more commonplace, I eventually graduated with a degree Electronic Arts at the Atlanta College of Art in 1997.
During that time, I interned for a CRPG by D.W. Bradley of Wizardry fame. That project allowed me to try a variety of positions like Modeling and Animation. I learned more about Photoshop during those years than I ever did in school! I then found an affinity for Level Design. It used all of my previous experience with art, film, programming, and even Lego building as a kid. This became my career path for the next 20 years.
Having lived in over a dozen places in the United States, you learn to see what’s unique and what’s the same no matter what town or social circles you’re in. Combine this with the variety of degrees I explored and you have a lot of perspectives you can apply to game worlds and non-linear experiences.
Being a Game Designer comes at a risk, though. While you learn how Art, Gameplay, Code, and Tech all have to work together, it becomes a specialized skillset. Artists and Programmers have more choices to move in and out of game development. A Designer is married to it.
I’ve been lucky enough to jump to bigger and deeper worlds for different creative teams. My goal as a World Designer is to make players forget they’re on a Mission or Quest. Have them poke the fictional world to see how it pokes back. Beckon them to explore the next corner or hilltop for new sights & wonders.
When you’re moving through the world, there are far better ways than staring at a map. Playing the “mini-map game” by constantly looking at your GPS map pulls you away from what’s exciting in the world around you! The best type of game is one that doesn’t give you a map scrap as a critical info hub.
Having said that, maps serve a greater purpose and mental itch. They usually involve mystery of the unknown and building anticipation to a destination. A dotted line to buried treasure! It’s also a picture of bookmarks and notes to think about. You can pocket it and share it with others, pointing out where all the good, bad, and ugly happened. It can be your trophy of memories to hang off the wall.
I always start a new gameworld by mapping the hotspots and pathways to and from them. What will make each region or realm diverse from one another? This quickly shifts into a 3D blockout to prove it has enticing vertical and horizontal folds.
When I worked on the sequel to The Suffering, I was given a linear route through each story and gameplay beat. The itch to make it a potential sandbox environment got the better of me. I ended up folding the linear route around on itself creating a loop. This also allowed me to let players glimpse at upcoming places and look back on past locations. The final game never took advantage of this loop, but it helped flex my sandbox instincts.
How important is realism?
Worlds shouldn’t be shackled with reality. The more it deviates from realism though, the more its outlandish rules need to be consistent and make sense. At the beginning of a project, I push the team to be 80-90% believable. As the project progresses when there’s now a solid believable foundation, let more exotic elements splash in. By the end, you can let half realism go out the window if it means that frosting on the game makes players sit up in their seat. Remind them it’s not the real world, it’s whatever it needs to be!
That’s one reason I look forward to DLCs… it’s a chance to take a well-seasoned team and try crazy ideas that wouldn’t originally be considered. At that point, the players are fans and can appreciate twists or even radical changes to experience the world in a new way.
I try to have 2-3 world state changes throughout the main game in an open world. It shakes up the way players move through now-familiar neighborhoods and regions. More importantly, I try to make those changes the player caused instead of a story beat. Then they can always be reminded their actions and power impact the world.
You have to fight for these as projects shift and get refactored to hit production dates.
On Mafia 3, we put a lot of effort building northern Downtown under the river’s water line. I wanted to flood that area and move through it with boats later in the story. Imagine ripping between two skyscrapers in a fan boat like Venice after a hurricane!
A mission on the other side of the city has players blow up a freighter in the shipyard. The original design was to have that slam and half submerge against the dock becoming a scar on the city. Story and production changes sadly bled both those ideas out of the final game.
The more outlandish a world is, the more open-minded players will be accepting what it is. The closer the world is to reality, the more scrutinized it’s going to be. Mafia 3 had to feel like it existed if you traveled there today. The team put a lot of effort constantly testing the reality of it. The Czech crew was amazing in this capacity.
Mafia 3 was also a chance to educate people, giving them a glimpse of what it was like to live back then. A ton of real-world history influenced what New Bordeaux became. I had a desire to surface all that knowledge as Tourist Spots the player could read when they encountered them, like a newspaper Wikipedia. Players could have finished the game with more real-world knowledge by exploring the city. Time ran out though and this was cut as well. The closest we got to this was the marketing team at 2K Games releasing New Bordeaux social media.
I was excited to see Ubisoft did something similar to Assassin’s Creed Origins “Discover Mode” to explore Egypt!
For anyone who hasn’t made an open world game, it can be a real eye-opener. Unlike a linear game, you’re building locations you expect or hope to be used, but they ultimately might not be. Since it’s a sandbox environment though, it’s extremely difficult to just cut a middle section out because no one’s using that space or wedge something into an existing space.
On Mafia 3, Donovan’s motel headquarters was originally north of Frisco Fields. Halfway through development Delray Hollow became the launching ground for the player’s first story arc after he meets Donovan. To make it easier to see Donovan, we moved his base-of-operations right into the middle of Delray Hollow. Those unique assets caused a ripple effect and closed off other planned locations as well as bumping the Community Church hideout further south. That wasn’t a simple task with so many details and technical links already layered on top. Talk to everyone on the team in an effort to minimize radical changes later.
Another hurdle was not realizing how time-consuming it was going to be creating and managing the spaces between quests. We originally estimated a quarter of the time to construct a district with half the people. It was actually far more challenging for gameplay as well. While these in-between spaces were less critical or hand-crafted like linear quests were, it still had to support competent combat. Having a player enter a back alley five different ways with any type of weapon and make it an exciting fight was difficult to balance!
A memorable world involves more than alluring physical spaces. It needs to work in concert with the gameplay and story. You have to know when places could be used and how many types of gameplay will use that space. Is it just for gunfights or will a car have to drive through it? A basic sandbox game loop involves adjusting player opportunities with what’s happened recently and what needs to happen soon.
Creating a large urban city is a lot more complicated than creating a large terrain environment. Filler areas in the wilderness can be random and organic. You can discover a lot of fun topographical pathways and points-of-interest through randomization. A city, however… everything’s constructed, everything’s there for a reason, even if it’s a filler building.
Mafia 3 had the advantage of being based on a real-world city. Everyone already had a knowledge of what a contemporary American city was. If someone wasn’t sure what details should appear in a backyard, they could easily Google images for hundreds of examples. A modern baseline, visual language already exists for things like traffic signs.
Creating a fictional, outlandish city has none of those advantages. Everything has to be created from the ground up and make sense within itself. You have to create a history of why something was made a certain way to make it believable. Without that, your world starts to feel like a game and not a place. The whole team needs to understand the baseline for imagining what takes place between concept art and core quests. There’s no Googling a team’s unique style and theme.
Building a massive world
If you have the procedural systems set up, it’s the best and quickest way to crank through dozens of broad scope layouts. Things like ESRI’s City Engine take time to set up though depending on how far from real cities you’re swinging from. If you’re creating a place that never existed before you have to first figure out what a basic city block looks like. Is it even a block? Only once you find how big or small the average building is, how wide the street should be, what ways the player will traverse this space, can you then multiply that outwards to draft the larger city and create diversity from that?
How many sandbox cities are out there today that don’t recreate historical towns, modern metropolises, or riff on what’s been seen in sci-fi film and games? Show me a large fully-realized, believable fantasy or outlandish city people want to go back to over and over, and you’ve found a place built by a lot of dedicated and passionate people who can’t just look at history books or out a window. They want their world to feel alive when you aren’t in it.
When creating a new world based on an existing place, find what things people remember about that location. Use those as tentpoles for landmarks and routes. It’ll squeeze out spaces people won’t be interested in, helping you compress your gameworld into something manageable.
Plan out the ways a city operates on a day-to-day basis. This includes how different classes and cultures mix into the city day and night. This is how to anchor diversity and create a backdrop of conflict for city life.
Early on, my general approach is to work closely with the Narrative and Quest team to understand what they want the city to do for them… the primary locations and ways the player will interact with the game. Also, understand or at least anticipate what tech features or limitations the team will have. Building the geometry for the city is only half of what goes into it. There’s audio sectors, lights, decoration count, people variety, vehicles types, and more that take bites out of the performance pie.
After you take all those requests and limitations, create a simple 3D mesh of the city. After a lot of rapid iteration, get the Leads and Directors to sign-off on the scope of the world. The basic world has to be proven before major quests make assumptions about it. Make sure there’s enough space between unique locations so they can stream in without hiccups. It can influence what happens before and after a quest. It can also help maintain the local theme by giving the team something to stand on and push up against.
Now the real game production can start. People can start focusing on regions and integrating their own ideas on top of this rough world outline. (I could write a whole chapter on world building production.)
On a side note, I’m excited to see Frogware’s The Sinking City. They’ve taken the time to set up procedure tech and rules for generating their supernatural 1920s city.
I saw Google Maps unlock their content for game content! It was one of the first things I thought of years ago when Google Earth let you travel through 3D cities. I’m excited to see what people will do!
On the flip side, it depends on what type of game will use the literal layout, whatever it’s skinned as. There’s a reason New Bordeaux in Mafia 3 is not an exact copy of New Orleans. There’s a lot of uninterest space in the real world. For example, I was anxious to roam the streets of LA Noire, only to realize it was too literal to Los Angeles and didn’t have the content to support anything but the main storyline.
Unless the world is literally a backdrop to gameplay like a flight sim or giant monster smashing, copies of cities will never be as good as something that’s hand-crafted. Small items within the gameworld like apartment furniture can be automated and randomized within a ruleset to reduce the amount of time it takes to construct them and fill them with interaction. It’s the macro layout and pacing from one arena and point-of-interest that needs to be judged by the human eye, playtested and adjusted over and over to make a great game experience.
It’s the early stages of a project when automated or copied content is fantastic! Using tools like Google Maps is a great way to get through several drafts where you think you know what you want, only to find something better with the next one.
These are the tent poles that influence the local themes, attract the player, and help spread and balance the world with diversity. It’s actually one of the easier things to do early on. You do it during the drafting stages I mentioned before. Mafia 3 had to balance the relative locations of known landmarks with unique quest ones.
Part of the drafting phase is building the right paths and routes from one landmark to another. In some ways, you can consider the primary roads and landmarks the backbone to a good world before any fancy parts get slapped on.
Mafia 3’s city went through 3 major layouts, followed by several dozen micro layout changes.
The world needs a history. It needs a story. What happened before the story started can greatly influence and inspire what’s happening during the story in the world. A world story is also something the team can lean on when they need to fill in gaps.
When developing a history, find ways to expose it to the player by showing how layers were built on top of other layers or left to neglected over time. An example of this in Mafia 3 is the small river island in the middle of the city. Despite being between the popular French Ward and Southdowns districts, inhabitants now overlook this place. At one time it was an active shipping and packaging hub before Tickfaw Harbor became the modern port. The buildings are now crumbling in the shadows of the modern highway spanning over it.
Like any game that has a player moving from one location to another, it can reveal an unspoken story. Who made that path? What’s that side trail for? Even if you get to a cabin and no one lives in it, who did and why did they leave? Almost every cabin in Mafia 3 had some loose background story to it.
A stronger challenge in passive storytelling is to keep things diverse. Don’t let every area be everything. The absence of something you saw frequently in a previous area can imply something about your current area.
In New Bordeaux, there’s an impoverished neighborhood in southwest Frisco Fields called Haven Ward. It’s where the slaves once lived when Frisco Fields was a plantation before converted to the middle-class suburb Hillcrest Estates. One member on the team was trying to make this struggling area attractive by planting some nice bushes and flowers along the road. I had to remind them Haven Ward is trying to find their identity in a district that doesn’t think African-American’s deserve to be equal. Let the neighborhood reflect their stark situation.
In contrast, Delray Hollow was another African-American community that was able to find pride, identity, and self-reliance. That’s partially why it was so far away from the center of New Bordeaux, away from the shadow of White Southern attitudes of the 1960s.
It’s challenging trying to figure out how big does the world need to be. Build to the expected content and pace the player going from Point A to B. There’s no clear formula and you don’t want to find yourself anchored to content that will never get used. Something too attractive with no substance will lure players only to be disappointed there’s nothing there – no rewards for time spent.
Mafia 3 had a lot of spaces never used. The original idea was to have 3 hideouts for every District. This was baked into each district, finding an appropriate narrative to fit a landmark building and enough space around it for streaming performance. Design shifted down to 2 hideouts, shelving the other for DLC. Its space in each district still had to be maintained throughout development, however. Near the end of the project, those DLC hideouts plans were dropped in favor of new story arc content. Some places were repurposed for the new DLC, while others were left untouched.
On the opposite end, you don’t want to be too small or you encounter performance problems later as quests and unique content smash up against one another. Something has to be cut or a lot of time has to be spent ripping apart the world to make more space in between.
So it’s always a gamble if you’re going to have too much or too little.
Again, this is a lot easier if it’s a terrain-based game and not a city-based game. I’m working on creating spotlight articles for each district in Mafia 3 detailing how it developed over time and things learned. After that, I’d like to take a real-world city like Berlin and flip it into an exciting and manageable game city layout.