Let's talk about the journey of creating...

Relics of Eldoria

Language: Lua
Engine: Love2D
Time to Complete: 3 Weeks

I envisioned creating a game that used a directional-based attack as the core gameplay mechanic when starting this game. Beyond that, all I had in mind was a dungeon where I could fight monsters. It was a fairly basic idea that grew the further I got into making the game. I began with the directional attacking by just getting a sword to move on the screen along the path of your press-down position to the release position of the player's mouse. I had just gone online and grabbed a couple of free random images to use as placeholders while I started getting some of the mechanics in place. Getting the basics for player attacking, enemy attacking, blocking, damage, health, experience, and levels in place were all priorities before worrying about the game's overall look.

After completing those, I wanted to take some time to get a feel for the game's looks. At first, I started getting some low poly 3D models I had purchased and using them to create 2D images for the game. Admittedly, I spent far too long making several weapons and backgrounds. Still, after all that work, I had only one enemy, a pixelated demon berserker for my enemies. It didn't fit the style, and I didn't have any low poly monsters, and purchasing them would have been too much for this assignment. I tried to find some free online but couldn't find enough that matched in art style to each other, let alone what I already had. This is when I chose to change gears a bit. I looked around for some AI art generators and started submitting prompt after prompt to generate enemies for my game that fit the theme and style. Bing Create absolutely came in and saved the day for me. It produced the first few enemies I thought looked great for the game. Seeing just how good these turned out I decided I would get AI to do all the artwork for the game so I could focus on programming the game. I knew I couldn't spend all my time on art if I wanted any chance at getting the game done in time. 

I have to admit, though, that I may have gotten a little carried away with the number of assets I added. I was so amazed by what AI could do that I spent four days generating images of weapons, backgrounds, enemies, and UI elements. For every image, I needed to scale it, remove backgrounds, and, for some, add shadows.

At this point, I had to stop and start working on my Game Jam Coffee Run. I was down four days and had to get back into the groove of programming this game again. I had enough editing images and was really happy with the progress of the game. However, I had spent so much time on the art that I felt the gameplay was lacking and got back on track.

The next big hurdles were finishing the fighting, looting, leveling, adding an inventory, shop, dungeon generation, and saving/loading. It felt like all I did was program, and it was a blast! Along the way, features were added, more art was needed, and time was slipping away.  Of course, at this point, I had to ask myself: what is a game without SFX and music? I knew I had a ton of sounds I had purchased, so I began listening to far too many similar sounds to get the best one for every action, button, and move in the game.  It's one of the few things I didn't get AI to do. The feeling of clearing your to-do list is always a great feeling, yet the game didn't feel finished.

Until now, the game was endless. Fight, get loot, level up, and repeat. No real purpose, no goal, which can be fun for some people, looking at all you idle game lovers. However, I wanted to give the game a story with an end and, of course, the ability to keep going if you wanted to after. So, keeping in the theme of AI doing the artwork and stories being art, I asked our friend Chat GPT to write a story for the game. It was here that Relics of Eldoria was born, and our hero, Aiden, breathed his first breath. I had to implement cutscenes into the game, get a nice-looking logo, get relics added in, and throw in a tutorial while I was at it. I had thought about just doing the voices myself, but again I was sure AI could do it faster and with more variety. I began the hunt for an AI generator for voices and stumbled across Replica Studios. It didn't take long; the characters all had unique voices and the story had come to life. With a few finishing touches and particles added in to try them out, the game was complete.

Or so I thought. I got some help from my QA team, also known as my family, and testing began. I was determined to show my class a game that not only looked pretty cool but also functioned. Is that too much to ask? Several bugs later, my sister absolutely killing me in my own game, and finally, we could get past a few hours of playing without it crashing. I really have to give it to my sister and mom. They played this game for a few days to help me find bugs. They are amazing.

Presentation day came, and Relics of Eldoria was complete.

Challenges...

The greatest challenge I faced while making this game was scope. When I started the game, I just had a simple idea, but the more I programmed, the more art I generated, and the bigger the game became. I had to tell myself to keep on track, focus on a clear goal, and push for it. Time was not on my side, and I could only do so many 3 or 4 am nights back to back and keep a sharp mind. Having the ability to cut out unnecessary work, stay focused, and have a plan with broken-down achievable goals is what got me through.

Though I was using AI to generate my art, being a perfectionist, at times, it can be easy to get caught up in getting things just right. In doing so, you sacrifice time, which is far more valuable in a case where time is very limited. Knowing this I set terms for myself, at what point is good good enough. Setting clearly defined expectations or defining what constitutes complete can help mitigate the time versus quality struggle.

As with any game, balancing is always a struggle. Small changes can have big effects. Balance can only be achieved through many iterations and plenty of testing. It's pretty hard to just luck out the first time you punch in numbers that seem right, and they just work. In the end, I feel there could be more balancing done, but I just ran out of time.

What I came away with...

Scope. Understanding it is a skill I think everyone will constantly improve upon. It's important to know just how long different tasks will take. Being able to say, "This task will take me this long," and being accurate, go a long way to helping producers roadmap properly. If possible, it gives you and your team the best chance to finish on time without the crunch.

AI is great for producing quick but high-quality art for solo game developers that don't have artistic abilities. After all, not everyone can claim the fabled title "unicorn."

An understanding of how to write prompts for AI to get better results. Be descriptive; be specific.

It is important to plan out enough time to properly test your game before releasing it. Having a good QA team is extremely valuable.