Heart & Core
Type: Group Project
Development: November 2022 - Present
Description: Heart and Core is a 2.5D Twin-Stick Brawler centering on two protagonists. Jean Jones, a human engineer, and K.A.I., her robot companion. Heart and Core will take these two characters across three different spaces (Junkyard, City Streets, & City Center) with the objective of fixing K.A.I. before it’s too late.
Purpose: Heart and Core serves as my team’s capstone game. My team, Chonky Raccoon Studios, consists of 17 producers, artists, programmers, and designers all working a minimum of 15 hours a week across 8 months with one single goal. Create a portfolio-worthy product that shows we are ready for the industry.
Tools Used: Unreal 5, Blueprints, Photoshop, Milanote, JIRA, Perforce, World Partition
Flythrough & Concept
With the ultimate objective of giving the entire team a portfolio piece, the conception of Heart and Core was far more collaborative than your average project. As Design Lead, I started the team on a blue skies approach, whereby each member who was willing would give their wish list of game genres, mechanics, art styles, etc. for the leads to comb over and find patterns.
After blue skies, a few patterns did emerge. I knew going off of blue skies that the team was centering on a 3rd-person single-player title with strong characters. These served as helpful first constraints for the design.
With these constraints in hand, I gathered the entire team for a 2-hour long meeting to find our heading. We needed a phrase for what our game could be. It could even be just a word, but it would need to set us, as a team, on a path and take into account the previously decided constraints.
What did we land on? An engineer and their mechanical pet.
Now with a rough heading, I begin to work closely with my art lead. Given how passionate the art team was about having a strong vision, I decided to break convention a bit and postpone explicit mechanic design in order to more fully establish theme, tone, and aesthetics. This meant art and design had to walk hand in hand. Design would hold a meeting, then consult with art. Art would hold a meeting, and then consult with design. This ping-pong method worked out quite well and landed us some further definitions of our game as well as constraints.
Theme: Letting Go
Tone: Dread
Aesthetics: Steampunk
New Protagonists: Girl and her mech friend
New perspective: 2.5D
Gameplay: Combat focused
From these new constraints, design began to concept a strategy game whereby K.A.I would be controlled via a command wheel and the player would be in full control of Jean Jones (JJ for short). This was the design we implemented into vertical slice, and you can see my thoughts below.
Screenshots
Full Playthrough
Where I Messed Up
I did it again, I made two games stacked on top of each other.
After the vertical slice, I realized our design had a fundamental flaw. There was far too much for the player to do at any one time and the two games we had made were constantly getting in the way of each other.
Was the player platforming as JJ? Very cool, now they’re completely ignoring K.A.I. because of cognitive overload.
Is the player commanding K.A.I.? Hey look JJ is completely still now because again there is far too much for the player to keep track of.
We ran headfirst into this issue for a variety of reasons but a couple of strong reasons stand out to me.
1. I let perfect be the enemy of good, and it cost us a lot of time.
2. I kept looking for ways to justify our two characters by always giving them something to do.
With the first reason, in order for K.A.I. to work in combat we needed not only the K.A.I. command interface to work flawlessly, but we also needed the entire structure of his time-based combat to work around him. With this, we as a team decided to build out this system responsibly so as to diminish overhead later on as we added features. What this also meant was that design would be unable to playtest this system until it was at least 80% there. This meant heading into vertical slice there was still an open question on the viability of the combat design. It was only just before the vertical slice that the answer to that question was a sound “no”.
Lesson learned: Prioritize MVPs and deal with the jank as we go, we’re more time sensitive than we think.
The second reason came from insecurity about our design. As a designer, you always want as many possibilities in a space as you can fit. You never want the player to feel out of options or even worse bored, but that can be a trap. In trying to have each character always have something active to do I didn’t keep the player in mind but instead the characters. This ended up creating two full players’ worth of mechanics and ended up overwhelming our players.
Lesson learned: Always contextualize your design to the player, do not run your design in circles without them.
So where are we heading currently?
Believe it or not, now you actively control both characters with each half of the controller (hello Brothers: A Tale of Two Sons). This new control direction, with paired down character actions, in 2.5D is serving as a much more promising play space for our game going forward. What’s even better, we can actually prototype this very quickly and find where it breaks.
Other Projects
UE5 Harry Potter VR Dark Ride Concept Gameplay Programmer/Level Designer
The Lava is Lava - FPS Arena Level Scripting/Design