Chubby Cheeks is a 2D platformer in which the player assumes the role of Chubby Cheeks, a rotund hamster who's on a quest for the fabled Golden Acorn.
The game was developed in the Torque engine by four students at the Guildhall at SMU. The total development time for Chubby Cheeks was ten weeks (fourty-eight person-hours per week), and was developed for both the PC and Xbox 360. The final product included a single tutorial level and a fully fleshed out adventure level.
Considering that this was our first time working together, were unfamiliar with the toolset, and had three other classes demanding our attention, this project went very well. This postmortem examines the development challenges we faced, and the triumphs we shared during the production of our first team game.
What Went Right
1. Team-wide investiture in Vision
We knew that the timeline was going to be unforgiving at the outset of the project. With a hyper-compressed timeline and other classwork to divide our attention, we knew that each member of the team needed a personal stake in the game’s vision. Otherwise, we risked losing team members to undermined motivation.
Fortunately, our game concept came about organically during the planning phase, and gave each team member an opportunity to have design input for the game. During a long brainstorming session with our artist and designer, we started down a long line of, “it would be funny if…” jokes. Our artist thought it would be funny if a hamster broke out of his cage and chomped his way to freedom, growing bigger and bigger as he went. We all latched onto the idea of a small, hungry hamster growing into a giant force to be reckoned with.
Coming to the conclusion that a Godzilla-like world destruction game was out of scope for the project, we started to ask each other how we could keep the core idea and make it work within our short timeline. Then we all said to each other, “What if he was just fat?” I started waddling around the room, showing how the hamster would bounce enemies with his massive belly. After a few minutes of hardcore laughter, we started to bounce ideas (forgive the pun) back and forth. From that point, we had fleshed out the game’s core concepts in less than an hour.
Excited as we were about the idea, I still had to tell the programmer, who was sick at home that day. He had pitched another game concept, and not knowing him that well, the team was nervous that he would shoot down our hamster concept entirely.
So I call him up and start explaining the idea. Eventually, I asked him, “So, what do you think?” After a long, stressful silence, a roar of laughter came back from the other end of the phone. “That’s hilarious! Let’s do it!” Immediately, he started offering his own design input and explained how he could make the main mechanic work in our engine. Hearing his excitement over the phone, I knew we had found our game. I notified the team that he was on board, and thus, Chubby Cheeks was born.
The virtue of such a small project is that each team member had design input on Chubby Cheeks. Each team member contributed to the overall vision of the game. This allowed each team member to become more invested in the project, and ultimately led to better team performance and attitude.
2. Sustained High Morale
Like any team, we had our good days and our bad days. Meeting milestones in other classes were major stressors, and often affected at least one team member’s morale. Fortunately, since each team member had a huge buy-in during the concept phase, the team as a whole remained invested in making Chubby Cheeks as fun and as complete as possible.
We always had at least one member of the team in high spirits that would boost the morale of the entire group. After each milestone, our artist sketched Chubby Cheeks dressed in the milestone’s ‘theme.’ At the end of Alpha, she drew Chubby Cheeks dressed in full Dovahkin-style armor to celebrate the release of Skyrim in November. After Beta, Chubby Cheeks sported a Jedi robe and a lightsaber while levitating an acorn to celebrate the upcoming release of The Old Republic.
These small morale boosters gave the team a positive attitude that got us through the tough times and injected tremendous amounts of fun into the development process
3. Prioritizing the Schedule
We received some extremely valuable advice at the beginning of the project: establish a single, primary mechanic and design the rest of the game to facilitate this mechanic.
Toward this effort, the team decided that Chubby Cheeks’ bouncing mechanic would be the only mechanic not native to our toolset. In other words, our programmer only had one mechanic to create, tweak, and polish.
In the planning phase, we thoroughly outlined how we wanted the bounce to work. Would the mechanic only serve as an attack? Would it affect how the hamster moves through the world? What challenges would it present from a design standpoint?
Once we had the answers to these questions, our programmer came up with a proof of tech for the bounce. That gave us great visibility as to what we could work with. The artist was able to see how the hamster would bounce, which allowed her to draw the animations with the hamster’s in-game movement in mind. The designer and I were able to immediately envision how the mechanic could be used in the game space, which allowed us to design a world that would reinforce the primary mechanic.
Overall, designing our game around this central mechanic kept the game elements consistent with each other, and was a big win for us from a development standpoint.
4. Cut Early, Cut Often
Early on in the development process, we had an art feature that the team was very excited about. In our original documentation for Chubby Cheeks, the hamster was going to grow in size as he ate acorns throughout the game. We received feedback from our professors that this was majorly over scoped from an art standpoint, and we decided to cut that feature. We had the option to keep that in the game, but thank goodness we took their advice. There is no way we would have been able to do three full sets of character animations in an eight-week period and actually produce a coherent gameplay experience.
5. Tester Feedback and Course Correction
Originally, Chubby Cheeks could only bounce off enemies to reach higher platforms. Once the enemies were killed, however, reaching that platform again was usually impossible. We worked tirelessly to fix the engine’s buggy spawner, but getting the feature to work like we wanted was proving extremely difficult.
During the Alpha review of Chubby Cheeks we received some extremely insightful feedback. One of our testers made us aware that he couldn’t reach a platform again because he had already killed the enemy. We made him aware that it was a known issue and that we were working to address it. As an a work around, he recommended that we convert one of our flower platforms into a bouncy flower that would propel Chubby Cheeks up into the air (much in the same way as the spring platforms in Sonic).
Since we were already post Alpha, implementing this change was risky. However, the team made the decision to move forward with the change, and the results in the next milestone were absolutely fantastic. The idea was brilliant, and the team learned the value of listening to your tester’s feedback and taking it seriously.
What Went Wrong
1. Clear Communication doesn’t equate to Clear Interpretation:
I believed that, as long as I facilitated clear lines of communication within the group, the communication issues would be largely mitigated. Although people acknowledged and understood ideas, the ideas were not all interpreted in the same way. The programmer’s interpretation of a hamster jump was very different from the artist’s interpretation of a hamster jump.
As such, the communication of ideas across the entire team was not perfect during the lifeline of the project. This ultimately led to more work/rework for the artist, which could have been avoided entirely if the information was disseminated in a fashion that was more consistent across disciplines.
2. Distractions with other class work
As previously mentioned, deliverables for other classes were major sources of stress for everyone on the team. It was very difficult to focus on Chubby Cheeks when you have a level design Alpha looming in the background.
On more than one occasion, each team had to utilize our team meeting time to finish work in other classes. This primarily took place toward the end of the project’s lifecycle. While all of the team’s work would get done eventually, not working on Chubby Cheeks led to more gaps in communication.
Furthermore, while we didn’t have to lose too much sleep on account of Chubby Cheeks, team members would have to stay awake all night to complete their other assignments. This often resulted in arriving to meetings late (if at all), and being absolutely dead tired during meeting times. Maintaining focus on Chubby Cheeks duties would have alleviated some of the resulting communication issues, but given the nature of the project, the issues were largely something to endure rather than issues that we could avoid.
3. Learning the Engine
Delivering a complete product in eight weeks is an aggressive goal. Doing so with an unfamiliar tool seemed almost impossible. Large amounts of time were devoted to simply researching how to implement basic gameplay functionality.
While we were ultimately able to implement each desired feature, the team was left with little time to polish certain features before we had to deliver the game.
4. First Team Project
This was the first game development project for three members of our four-person team. While I believe we did very well considering this situation, there were growing pains for us as a team.
Learning how each person worked and adjusting the team dynamic to complement those styles was a very big challenge for me as a leader. Furthermore, learning each members’ preferred communication style was also a big lesson. While I tend to prefer verbal communication, other members of the team preferred written feedback. This isn’t an earthshattering revelation in the study of group behavior, but it did offer tremendous insight as to how preferred styles of communication affect a person’s work process.
5. Location, Location, Location
Each team was assigned a room for their team meetings. Our programmer brought in a single cup Keurig machine, our artist offered her decorative touch, and our designer brought in a mini-fridge. We made that small corner room into a cozy little home for the eight-week production process.
However, our building has very thick walls, and our room happened to be at the worst location for the facility’s internet connection. While it may sound a bit ridiculous, this shoddy internet connection had a huge effect on team morale. Nothing was more demoralizing than a solid three-hour work session, only to have your internet crap out during a commit to the repository. This usually resulted in errors that took time to resolve and prevented other team mates from attempting to commit their work.
We worked around this by taking our laptops down the hall to where the connection was better during commits. Still, it was a small problem that had a large impact on our workflow.
What We Learned – Conclusion
Chubby Cheeks was our first team-based game development project. This experience not only taught us valuable lessons in game design, but also about team dynamics and communication.
We will definitely make more mistakes in the future, but I’m confident that we won’t repeating these mistakes in our next project.