Hello again!

As the semester comes to a close, so does the development of Gobby’s Farm. If you’re new, Gobby’s Farm is a student project developed in the Advanced Production class at Chico State Universities Computer Animation and Game Development program. It was led by our amazing designer Kaeli and produced by myself, Nolan. Along for the ride we had 11 developers creating amazing work that you can see only if you download and play! With that being said let’s hop right into what went right, wrong, and everything in between.

Throughout the 16 week production, I assigned a total of 656 points, of which 607 were completed. Wow those are way bigger numbers than I am used to and I am so proud of the entire team for making this happen. This left us with a final velocity of 86.7 average points completed per sprint. It’s almost sad to see that train stop rolling.

What Went Wrong?

Let’s rip the band-aid off. We definitely didn’t start strong on the communication front. Regular in-person meetings started way too late, and that meant people were kind of doing their own thing without really knowing what the full game even was. Face-to-face talks are an AGILE value for a reason, and we learned that one the hard way. It got better later, but it made for a bit of a rough start.

Card priorities also got a little murky after the prototype phase. Audio, animations, and some level design got left in limbo because I leaned too much on devs already knowing what to do. That left some teammates stuck in the void for the first couple sprints. As the producer that was a kind of embarrassing time but I learned quickly how to overcome it… put the work in!

Also, cursed trees. Yes, cursed. We had models where trees were stacked in a group, like 3 trees inside each other, which meant the poly count was too high for many computers to handle. Same with rocks. This was a problem because it limited what kind of playtesters we could get. And almost ruined the experience for most, including myself. Thankfully it was a silly and fixable issue.

Build nights. Ohhhh build nights. A build night coming up meant catching up on your sleep the night before. Huge shoutout to the devs who stayed up late to push out great builds despite Unity randomly deciding to break stuff. Special shoutout to Carlos who would fix bugs until his morning class. I bought him a couple Red Bulls for his troubles. Jokes aside, I truly believe he is the best programmer in the program right now. His attitude, skills, and workhorse mentality truly brought life and fun to our game and our team. Hero.

We also had props missing because implementation cards didn’t exist early on. We’d model, UV, and texture, but sometimes stuff just didn’t make it in. I learned the hard way that when it came time for our level designers to put the props in the game, I’d have to write several more cards and give it to them immediately, rather than having them ready. Work smarter, not harder.

Oh, and the mazes… Look, no one likes being dropped into a maze with zero warning. We had too many and they weren’t fun. Lesson learned, stick to the vision, and don't overdesign for the sake of "challenge." This was brought to our attention when we had a mid-project meeting with our professors. I think I remember one of their feedback was “What the f— is up with the mazes”. Noted.

What Went Right?

The art style? Incredible. Gobby’s Farm had a fantasy storybook look that really stood out. Colorful, inviting, and full of charm. The visuals Kaeli prepared really helped us make sense of mechanics like Gobby’s hands and made it fun to build enemies and props around that theme.

Also, the story and cutscenes were fully implemented! Even with tight deadlines, we made intro and outro cinematics (with two endings!) and even in-between week scenes to keep the player up with the story. We got voice acting from friends and family, and tied everything together. Was it overambitious? Maybe for a seven sprint student project, but we took it head on and never looked back. That's what I love about the team.

Pre-production was a dream. Kaeli knocked it all out in a weekend. We had spreadsheets for models, UI/UX, animations… everything. Before we had access to Jira I made a google doc with over 300 user stories ready to go for sprint 1 kickoff. The team jumped into production with a full backlog, reference images, and GDD in hand, and it made my job so much easier. Thanks Kaeli :)

The prototype and game loop came together fast. The core mechanic was clear: collect enough produce in a week to progress. We had inventory limits, a timer, and even a hidden ending mechanic. We axed an idea about sniffing poison produce and even redesigned the whole map, but it all worked out.

The team chemistry was off the charts. Good vibes and friendly attitudes made it easier to have real conversations during critiques and kickoffs. People gave feedback, took feedback, and worked together. This made the whole semester enjoyable and I feel so fortunate to have a great team.

Lastly, our velocity increased 5 sprints in a row! We found our rhythm and absolutely crushed it.

What Would I Do Differently?

First off, I’d make sure people are reading the game design doc. Kaeli spent an entire weekend on it carefully making sure everything in the game is accounted for. It really could answer any question! It was updated and organized, but if no one reads it, what’s the point?

More in-person meetings. Being able to look at someone’s screen and immediately make a call helped us late-game, but it should’ve been there from the start. Same for team playtests, bugs that we caught in Sprint 7 should’ve been seen way earlier.

I’d also fix Sprint 1-3 pacing. Level designers didn’t get much work early on because we were still waiting on models. This came down to a lack of experience on my part. I should have put more work in to give the level designers more to do. And we definitely should’ve given better contact info up front. Not everyone checks Discord.

Lastly, verification needs to be more than a Jira screenshot. We should’ve jumped into Unity more often to actually see if stuff was working. This led to more bugs and more fixes later in the project.

What Would I Do Again?

Like I’ve already said, the pre-production was amazing. Absolutely doing that again. Having a big organized backlog ready before Sprint 1 was crucial to a smooth start as well.

We had models coming in fast, feedback from playtests, and ideas clearly drawn out for everyone. Knowing what our teammates could and couldn’t do helped a ton in assigning tasks, and we had enough flexibility to help each other out when needed.

Short meetings worked too. Even though we didn’t have a lot, they were focused, and screen-sharing led to a lot of spontaneous critique and improvement.

What Did I Learn?

I learned a lot about myself this semester.

Managing a big team, balancing moving parts, solving last-minute problems, it was chaotic, but it felt right. Throughout college I’ve found myself comfortable leading and managing people. It’s fun, and I think I feed off of the interpersonal connections I make along the way. I’ve been doing it outside of my game dev projects too and the most rewarding part is seeing something being built from nothing and ultimately a finished product that I am proud of. Trying something new like this was scary, but having a strong vision and people that believed in it was honestly one of the most rewarding things I’ve done.

So yeah, Gobby’s Farm is done, but I’m walking away with a lot more confidence, experience, and appreciation for the kind of work we do.

Thanks for reading, now download and play Gobby’s Farm!

For Gobby!