3C Design for large scale traversal systems
The freedom and fun of moving through large spaces
As technology has grow to be the super fast, super efficient version of today, so have our game worlds. It’s possible to render billions of triangles, render lighting, render geometry and collisions, have hundreds of interactive objects, and showcase a variety of visual and functional materials.
While worlds inside video games are getting larger and larger, their density varies based on the genre and themes. The increase in quality of art, level design, and visual storytelling now requires level designers to create sparse and empty locations where players can take in the beauty of the worlds they inhabit. The empty, barren spaces of Zelda: Breath of the Wild have a function in the world — to create a sense of mystery and exploration, or to simple bring some spatial serenity. The physical space of Red Dead Redemption 2 may be sparse to showcase the sense of wilderness yet untouched by civilisation, but the world is anything but sparse. Both these games are sandboxes with a highly complex ruleset in which the players can create their own joy, which is supplemented by the overarching level design of the world. Every mountain, valley, cliff, street, village, or structure can be explored. The sheer scale of the world space to accommodate all these locations, vistas, venues, and sets is tremendous, rivaling a real world experience of hiking or urban exploration.
Besides the world itself, there are now many nuances to imitate real world situations — unique terrains and geological difficulties. There are many uniquely functional hindrances in today’s games, including slippery ice, blinding deserts, insurmountable mountains, etc. Overcoming geological challenges in itself has become game features.
Too much world design, where are the 3Cs?
3Cs, or known as Camera, Character, Controller are the representation of the player in the game space. Camera is what the player sees, Controller is the player’s input, and Character is the physical vessel that represents the player.
Together, these three dictate how the player will interact with the game space, game world, and game mechanics. So what are some examples of the 3Cs?
The way the camera tracks Ori, what it highlights, and what it brings to focus as its subject is carefully crafted. The things she does based on player input is the controller, and what she can do is the character. The character includes her physical dimension, how much she will jump, how much she will slip, slide, skate, gleefully hop, and move around in her space. The same can be said for Mario, who does most of the same basic actions which feel immensely different due to responsiveness, momentum, acceleration, and effect.
3Cs for the new game worlds
Most of the game worlds are now three dimensional, and huge. Entire cities, islands, continents, and even all of outer space can be traversed in them. Yet moment to moment action also includes more intimate motion that can happen on a street, in a room, or in the social vicinity of other characters. So how do these games cope between the requirements of such changing scales?
The same character and same controller must operate in spaces of vastly different scales. Arthur Morgan is a representative of the old wild cowboys- bulky physiques, laborious movement. The way he reacts to the controller lends to the fantasy, but it is simply not possible to use Arthur Morgan to traverse the wilderness of the world. Hence, the players get access to horses, coaches, and trains. All of these are built to cover vast distances and hence control much differently than Arthur.
Covering the huge expanse demands a lot of time being spent on horseback, yet being on horseback is not the best of Red Dead 2. And hence there have been many complaints about the world design, despite the world not being at fault. Should an activity dominate as much time as riding horses in Red Dead 2, they should be designed for inherent fun and polished to a sheen.
The games I really want to talk about
Spider-Man 2018 and Batman Arkham City.
Despite videogamedunkey ripping on how overused the term “feel like Spider-Man” is, the term indeed stands true. Swinging through the dense structures of New York’s skyscrapers indeed makes us “feel like Spider-Man” more than grappling and gliding our way through Arkham City made us “feel like Batman”.
Batman : Arkham City
Batman was never made for large scale traversal. His first outing under Rocksteady was a fairly linear, stealth game called Arkham Asylum, where he took advantage of ducts and overhead vantage points. He used his grappling gun to access spaces outside human reaches, and glided across architectural gorges to reach the other side. It was believable when the game was situated in the small and tight Arkham Asylum.
When Batman moved on to Arkham City, it was a different story. The same grappling hook was now used to traverse across a gigantic city, the gliding was used to cover distances spanning multiple city blocks. No Batman movie or comic is dominantly about Batman covering huge distances without his Bat-mobile, yet Arkham Asylum’s most dominant activity is traversing the city to reach locations. Yes, it is more dominant that the actual stealth, the star of the gameplay. Let us analyse the 3Cs of Arkham City and understand why it is not the best example when it comes to large scale traversal.
Batman is a huge character. He is bulky, muscular, and wears a gigantic bat costume with pointy ears. He has tons of gadgets and has the grace of a monster truck (except when he is fighting). Also he is an actually huge character, taking a 3rd of the screen.
Batman has a grappling hook that he can shoot towards a grappling point. The hooks sinks in at such a point and drags him towards it. Batman can cancel the motion at any point and drop into a free fall, but he cannot change direction.
He is slightly more controllable when gliding, where he can change direction and pull a few maneuvers like diving to build up momentum before unfurling his cape and gaining a lot of height. He can also shoot his grapple while gliding, and use the momentum of the grapple gun to launch himself over huge distances. The launch is a huge affordance in a city where height differences between streets and rooftops is a lot.
Changing direction is a problem for Batman. Due to his rigid grappling hook, taking advantage of using himself as a pendulum across corners is impossible. Cornering using a mass and rope is a massive miss in a game all about ropes, mass, and grappling hooks. It constantly requires Batman to glide to navigate corners.
Batman is pretty snappy when it comes to reacting to the controller, automatically locating and snapping to the edges of architecture. He travels mostly in straight lines and normal projectile curves when grappling. Batman actually shines when it comes to gliding as he has a much favourable sensitivity than compared to his on foot controls. It is possible to navigate tight spaces when gliding, and navigate tight vertical spaces using diving and cape lift-offs.
Of all the 3Cs, this one is the most bland in the Arkham Games. Batman’s motion is pretty independent of the camera, but the camera is very dependent on the character. Hence it difficult to align the character to where the player wants to look, but it automatically (and sometimes frustratingly) aligns to where the character is facing. It also does not use any sleek camera tricks that are used when capturing high velocity motion.
For a game that is dominantly large scale traversal, it fails to create and sense of thrill or excitement. At most, some motion lines are present and a change of distance between the character and camera. It lacks any dramatic change in FOV to bring focus and tunnel vision, it does not have auto pans and tilts to bring a sense of verticality, and lack dynamic distance change to bring a sense of acceleration. All pans and tilts are manually done as players use the controls to see anything besides where the character is heading. The only change in FOV occurs when Batman is using his grappling hook, not when he is gliding.
Unlike Batman, Spider-Man was very much made for large scale traversal. As this is the first entry in the series, Spider-Man was made from ground up with large scale traversal in mind. And unlike Batman, much of other Spider-Man media covers his relationship with a huge city like New York and how he manages the commute (with or without superpowers).
New York is not like Gotham though. While both feature tall skyscrapers, Gotham has a mix of building of various heights in the same place. New York on the other hand clubs tall building together, medium sized building together, and flatter architecture together. This eliminates any steps, ladders, and terraces in an area. When Spider-Man moves through a burrow/ block of New York, he travels through a uniform height of architecture. Which means the preferred height and scale of verticality does not change frequently and the player can focus on the pure horizontal motion of Spider-Man.
While Batman’s motion originates from aiming and shooting a grapple hook at different distances and achieving verticality to launch himself, Spider-Man shows a more rhythmic horizontal motion where shooting webs is the rhythm rather than generating the force for launch and motion. Batman does not swing, he generates force so he can glide. Spider-Man swings, which introduces rhythm to the motion. This creates the first and a fundamental difference between the two games’ traversal and 3Cs. While I have no experience in Level Design, as a Gameplay Designer it is exceptionally important for the players to be able to experience spaces and volumes in a fun way.
Spider-Man is a slim, lean, agile, and an athletic character. A huge part of Spider-Man is not only reaching from A to B, but the style in which he does it. Complex, athletic movements coupled with tight gymnastics is how Spider-Man has always moved, and the game encapsulates it very well.
Much like Batman, Spider Man can launch himself from wherever he stands. But unlike Batman, he is free to choose and change his direction right from the start.
Spider-Man actually achieves different horizontal velocities while swinging at different vertical heights. While this completely defies physics, this gives players a reason to vary their own sense of verticality. It is quite relaxed and leisurely at higher altitudes, which actually enhances the amazing view. Whereas the speed gained is amazing and thrilling at lower altitudes, which is massively enhanced by the challenge of dodging incoming traffic and slight loss of control due to the increased velocity.
Spider-Man is much more reactive to the controller on his feet and in the air. While the game borrows a lot from the snappiness of the Arkham series, there is much more free form control when Spider-Man is swinging. This allows even better mobility and allows organic, physical dexterity to come into play.
One thing this game excels at is allowing Spider-Man to travel in style. Not only does he do a number of animations, but players can also do stunts with Spider-Man. This creates an intrinsic challenge where not only reaching from point A to B matters, but the style in which one reaches. It is largely optional, but for an activity that is so dominant of the player’s time and attention, it adds some change and inherent challenges to spice it up.
The camera is one of the best aspects of Spider-Man. It does not only enable the fluid and well animated character, it enhances everything. It creates a sense of motion, momentum, acceleration, and change while anchoring it to a relevant three-dimensional context. There are many layers to the camera that individually react to the controller, so let’s tackle them one by one.
Camera and controller
The reason why Spider-Man feel so obedient is that he reacts to player intention more than Batman. The left analog stick is solely dedicated to controlling Spider-Man and right analog stick is dedicated to controlling the camera. But Spider-Man actually aligns himself to the camera, so it’s quite simple to steer him. It’s as easy as pointing the camera to where the player wants to be. So directing the character directly causes the camera to align, and directing the camera causes the character to align. Access to both these ways allows the system to be approachable by many different kinds of players. It also offers a chance for more experienced players seeking to utilise the system (I personally used the left stick for coarse alignment and right stick for fine adjustment)
With such a simple yet doubly faceted approach to camera, it’s no surprise that the cinematic scope of the camera is not lost regardless of Spider-Man existing and being used in a video games.
Achieving the feel of beautiful motion
While there are many nuances, this is the core rhythm of the camera movement. The camera has a slight spring to amplify or reduce the distance, creating a sense of acceleration or deceleration. Meanwhile a very subtle and periodic pan in the vertical direction shows off the rooftops above and streets below to create awe and thrill alternately. Here are two stills from the game capturing the same feel.
The change of direction relies on one of the most beautiful implementations of a dutch angle, that enhances the centrifugal motion. It lets Spider-Man cut away ever so slightly to create a sense of intent and reaction. It further enhances the sense of thrill by adding a curving motion to the camera.
Why is Spider-Man’s motion (or its perception) so heavily reliant on the camera?
Because it is not realistic. Batman, as a character, has much more realistic depiction of mass. Batman actually has generate torque using physically (somewhat) realistic centrifugal force. This disables him from making sharp and acute turns to orient himself. On the other hand, Spider-Man defies many properties of physics. He unrealistically defies principles of torque, generating large centrifugal force despite having a very small radius (where r in mr² is unrealistically small). Yet the beauty of the motion and the way the camera presents it to us allows suspension of disbelief, and allows Spider-Man to take advantage of a better reacting controller.
What should we learn from Spider-Man 2018?
Due to the scale of the worlds, traversing these worlds has become a dominant activity of the game loops. While the usual “reach location -> complete objective -> reach new location” loop for open world games has kept its form but changed its function for players. “Reaching locations” has become one of the most time consuming activities and therefore competes with “complete objective” as the dominant activity.
What Spider-Man teaches us that this activity can no longer be overlooked and should be given as much care and attention as the more meatier mechanics of the game. It has to not just be functional anymore, to allow players to travel from point A to point B, but also be carefully crafted for a different kind of fun — passive fun. While it is a dominant activity, it will not be on the forefront of players’ minds when they are more focused on reaching the location to probably shoot, hack-n-slash, or solve a puzzle. But as games achieve a sense of wonder to invite exploration, large scale traversal has become one of the biggest priorities.
Some other examples
Self sufficient characters are not the only options. While both these superheroes can inhibit small and large spaces using the same 3Cs, characters like Arthur Morgan rely on vehicles. His self is not sufficient for large scale travel so he gets a horse. Link from Zelda, Geralt from Witcher, Aloy from Horizon Zero Dawn, Arthur from Red Dead 2, and Snake from MGSV all rely on their mounts. Yet there are a great many differences in how these animals control.
Horse riding in Witcher 3 is one of the clumsiest implementation, despite the game being a pinnacle in many categories. The horse has too many animations running when it turns, thereby increasing the turning radius to that of a limousine. While it tilts and aligns well on long distances, the animal has a tendency to collide against level design geometry. This becomes frustrating once we go to the denser and more detailed sections of the Witcher 3’s world.
Meanwhile, all the psudo dino-horses from Horizon Zero Dawn were carefully crafted to create no nuisance while riding through the super dense world. The world is full of forest areas that have dense allocation of trees and rocks that would create a navigational nuisance if it were anything like the Witcher 3. Instead there are tricks under the hood that allow for some auto navigation and ease of life, which allow players to focus on other things like shooting while riding. After all, shooting on horseback becomes a very dominant activity when traversing between mission locations in a huge world. The character and controller are both very accommodating of the player and allow them to interact with the world in other ways than just to focus on the traversal, thereby reducing the monotonous nature of the activity.
Such careful emphasis on a traversal system that actually de-escalates its importance in the player’s mind is a starkly different approach than Spider-Man where they allow the traversal system to shine on its own.
While horses are becoming commonplace (along with bows), one of the greatest pioneers of large, open world games actually used a more modern vehicle — cars. GTA, Watch Dogs, Far Cry, and anything set in a modern time line all feature cars. Some open world games are made just for cars (such as Forza and Crew). Even Batman Arkham Knight, successor to Arkham City, features the Bat-Mobile as a dominant form of traversal. Some of these games also feature boats, jet-skis, planes, helicopters, etc. But the two games that I want to bring to light in this category are Saints Row 4 and Infamous Second Son.
Saints Row 4 is the world from GTA but on performance steroids. While it doesn’t have the largest area, it does have some height to it. The game allows players to traverse using traditional cars, or even air space vehicles that can switch between hover mode and propulsion mode. This allows many ways and speeds at which the city can be experienced and destruction can be caused with.
But the game goes one step further, it allows a self sufficient character who can travel without any vehicle. Naturally this level design is not the best at conforming player abilities, but the sheer amount of ways to traverse make it joy and give player agency to spice up their own, personal experiences.
Infamous Second Son uses its variety of transportation differently. While open spaces can be traversed in any way, most architectures come with one or two ability specific hook points. This creates a puzzle like approach to climbing and traversing massive buildings, ranging from sprouting wings and gliding, hovering on a smoke, or being a pure force of electricity. Various building and spaces demand the players to use these different abilities and their different traversal mechanics to keep variation in an otherwise monotonous task of moving between point A and B.
The way a singular activity of moving can be shaken up by giving tools to play with and demanding different tools for different challenges can really make traversal fun and challenging.
What if developers don’t have the time or ability achieve all this?
With new technology and 3D pipelines, it has become easier to create big worlds than polish and perfect systemic features. While many games have open world, focusing so much on a single traversal system may not be possible for everyone. Fast Travel is still one of the most used and ubiquitously present system. But that too, can be fun.
With Namco’s patent on loading screen games having expired, it may be possible to see minigames in loading screens when the level design chunks are loading. Or another approach is to make a transition through load and wait times by not disconnecting the players from the immersion. Spider-Man does that by showing a small snippet of urban NYC lifestyle, such as Spider-Man travelling in the metro via a cutscene.
By showing a snippet instead of a bland loading screen, the immersion is maintained and disconnect from the fantasy can be avoided. The snippet is used to show a slice-of-life in scope of the larger world elements, which in fact emphasises on the believability of the world. Storytelling can certainly be an option if interactivity cannot be achieved.
What is the future for this component of games?
Like many things in game, the future could be some version of Virtual Reality. VR has proven to be a difficult challenge, from a design and technical point of view. The biggest problem with the human body is that if the sense of motion that a body experiences does not match with what the eyes perceive, we get nauseous. So covering large distances rapidly while keeping the body still in a room is nearly impossible right now. There are many various haptics and ergonomically viable devices that can bridge the gap between the virtual and physical realities.
But withing what is possible and what can be cheap, the methods of traversal in VR range from continuous motion like regular video games, teleporting, and dashing. With continuous movement being the most immersive but being the most nauseating, the remaining options cannot fully harness the immersion potential of VR.
Context is the most important thing in physics, where a players’ eyes need to maintain the surrounding at all time that matches what the body is experiencing. But teleporting breaks total immersion as it makes traversal instantaneous. If continuous movement gives the most context but causes nausea, then the best option that has been so far decided is in hops or dashes. This may break immersion slightly but certainly maintains context better than teleporting.
Bring this to large scale traversal, at the current it seems very difficult. Various options involving maps have been explored, such as point and click to reach. Perhaps the future lies in non-Euclidean geometry. No matter what approach, VR has the highest potential to let players find their own fun.
Humans have a tendency of seeking fun in monotonous activities. We balance on rail tracks, we avoid cracks in the pavement while happily skipping over tiles. Sometimes we even seek company so we can chat while we walk to our destinations. It’s just human nature to seek fun and variety during activities that are largely one dimensional in complexity.
As designers, it is our duty to decipher behaviour, tendencies, and motivations so that we can best create fun experiences out of the oridnary.