A game about discovery in an Antarctic research facility
For these final images, much of my responsibilities were in creating many of the props used to detail the world. I also spent some time set dressing and adjusting lighting.
One of the major criticisms of the Radiant Dark BETA was that the puzzles, although the highlight of the experience, didn't feel like they had context. People didn't really understand what they were doing where they were, or how they were being used. The other issue was that solving these puzzles wasn't giving a good enough sense of progression.
That was my main focus for the redesign of the Inua facility. I went with a hub-and-spoke layout. This time, the puzzles were all in a central hub location, and the mechanism that swapped puzzles was unreliable, forcing the player to adventure out into the wings to get it running again.
We hadn't even completed greyboxing this space before Eli and I realized that this sketch to the left was NOT a very good space for the art we wanted. We decided to make the level more organically, creating a simple layout in our heads, and building the layout in a greybox, skipping the drawing. Below is what we came up with excluding some of the extra rooms we added as stretch goals.
One of my responsibilities on Radiant Dark was creating many of the prop kits. This includes consoles, mining equipment, office stuff, and a handful of small, detail-adding props such as vents.
When concepting began, we knew there were a few constraints on the model that we wanted to meet. We wanted it to feel more like a tool than a gun, it needed to have a way to vent heat, we needed a way to store the four orbs it fires, we needed room for expansion, and we wanted the UI to be displayed on the weapon, not the screen. Below are some really rough concepts I made to block out the shapes.
One of the more exciting things we did this quarter was creating an Art Book. Though time constraints lead to incomplete art being in the book, Wren King made an awesome physical book that the team can hold onto. It includes commentary from the team on their work, progression images, and helps to build the story of Radiant Dark. You can view the pages as a PDF below. I highly suggest taking a look, Wren did an excellent job with it.
Radiant Dark BETA
The BETA version of Radiant Dark acted as a proof of concept. It had a different art style, but was a full game start-to-finish containing roughly 5 hours of gameplay to new players with a massive facility to explore and 27 puzzles to solve.
We decided later to scrap this and go for a less experimental artstyle. We also removed the energy feature. Though it was more unique than what we ended with, and I think it has more potential to be interesting and fun, it was causing a technical challenge with all lighting needing to be dynamic and a gameplay problem with the ability being confusing; the team simply could not step up to that challenge and make it work. None-the-less, the BETA version represents the gameplay we wanted more faithfully, while the version above represents the AAA quality art we were aiming for.
After ten weeks, about halfway through the development cycle for Radiant Dark, we've come a long way. We set out to make a game that could be sold, and I think we got very close to that goal. It takes an experienced player about 90 minutes to complete the game and it can take a new player upwards of 5 hours, making it a full experience. The game has a beginning and end, a fleshed-out story, AI, music, few bugs, and has gorgeous art direction.
Though I wore many hats during the development of Radiant Dark, my most important was that of the level designer. Below I will go into more detail about the level design process and some of the changes that were made along the way.
I realized that it didn't matter how cool the puzzles were if the players would get bored before even arriving at them. So I went back to the drawing board and came up with a new plan that was smaller, more compact, and more vertical to keep things interesting. I also gathered significantly more reference material to craft from.
In addition to combining all the puzzle and buildings into one, large interior space, we also made the intro very short so the player only has a few moments in the exterior before they are placed into the game; previously it took nearly five minutes to get into the first building and really begin playing.
This new layout was not without its numerous issues, however. Many players still felt lost in the, still quite large, facility. Over the next weeks I would make many cuts to the scale; this generally involved removing extra rooms and hallways, blocking areas that weren't needed from the player, compacted some of the larger spaces, and made a big decision to make the game completely linear whereas before the path the player took through the game was mostly determined by which puzzles they could understand and solve at the time.
Among those changes, I also made a few changes that allowed the player to more easily navigate and know where to go. I blocked off sections of the map with colored key cards and doors, I created puzzle chains for more linear progression which helped the puzzles give a better feeling of accomplishment, I placed in emergency doors and lights which opened or lit up to guide the player to the next set of puzzles once they completed a chain, I created aurora arrows which pointed to the next puzzle to make sure the player at least knew the general direction they had to take to make progress, and I opened up sections of the map with windows and balconies to allow for foreshadowing.
None of this would have been possible without the amazing architecture kit made by Eli Gershenfeld and the gorgeous textures created by Wren King. Links to their websites are below.
Though I am a level designer and spent a majority of my time designing the level, I spent a lot of time working on the puzzle as well. The puzzles really are what makes this game unique and helps to tie it all together. The design philosophy here is adapted from Jonathan Blow. We wanted the player to know when they've encountered a puzzle, know exactly what must happen to solve it, and think about the solution, and once they know the solution it should be easy to solve.
The puzzle has a few basic rules that are as follows:
1) pieces can move around rings and can be pushed inwards towards the center on the tracks.
2) The pieces cannot be pushed outwards.
3) The center must have an equal number of blue and red pieces.
4) If pieces of the same color touch, they bounce away.
5) If pieces of opposite color touch they lock in place.
To aid in the creation of the puzzles as the programming was being done, I made a Photoshop document that allowed me to make the puzzle in a modular fashion to be tested. At this time it should be noted that we thought having the puzzles based off atoms would be very cool and help to push the science theme of the game.
In general, I hoped to teach the player something about the puzzles with each and every one.
I want to say that this only somewhat worked. Even now the lessons of the puzzles aren't perfect and about 70% of the test players were confused by the puzzles. I have two solutions for this problem.
1) I plan to increase the number of puzzles and explore each element of them more thoroughly.
2) I plan to make the lessons more obvious so that the puzzle cannot be solved until the player has learned the lesson.
The puzzles came out great. Not only are they satisfying to watch and solve, but they look gorgeous and are completely functional. Many of them have lots of opportunities for emergent game play with play testers finding solutions through ways I never thought of. Below you can see some of the puzzles in-game.
Those basic rules alone would make it impossible to create the 27 different puzzles found in the game, so to keep things interesting Benton, Eli, and I came up with many ideas for obstacles to put on the tracks. We came up with the following:
1) Wall: An obstacle that would block the pieces and deactivate them
2) Gate: An obstacle that would change the color of the piece that travels through it or bounces off the side of it.
3) Bouncer: An obstacle that forces the pieces to move in a certain direction at intersections, this was the only way to make a piece move to outer rings.
These three obstacles were the only ones to make it into the game, unfortunately, due to time and scale constraints, but there were many other ideas for obstacles:
4) Magnet: An obstacle that would push a piece to the opposite side of the track and move in tandem with it until both were moved to different rings.
5) Fences: An obstacle that would stop the player from entering certain parts of the puzzle.
6) Selector: A variation on the gate that only allowed a certain color through it.
7) Delay Pad: An obstacle that would stop the puzzle pieces for a certain amount of time.
Obstacles were not the only way we thought of to create variety in the puzzles. We also thought about rotating the rings, placing the puzzles on walls so only the bottom could be interacted with, making the pieces only be movable on a rhythm, and much more.
These puzzles could not have been made without the amazing hard work of our systems designer Chris Schickler. To learn more about the puzzle and how it works, I implore that you check out his website.
The two others that contributed the most to the design of the puzzles are Eli, whom I've already linked you to, and Benton Pellet, who was always there to disagree with all of my ideas and improve the game overall because of it. He can be found here. Eli also created the fantastic artwork for the puzzles and without his work they wouldn't be so cool to look at.
When I was waiting on player feedback I found myself creating many of the props that can be seen around the facility. The art style dictated by our art director was similar to Ashen, Absolver, and The Witness. We wanted a style that was realistic and captured all of the detail and function of the real life counterpart but with all of the excess detail removed.