PuzzleBoxEscape

1) Overview of the Game

My final VR game is a simple puzzle-based room where the player completes three physical puzzles to unlock a door. Each puzzle involves picking up pieces and placing them in the correct spot, and once they’re close enough, they snap into place. Each puzzle is personal to me, involving photos of me, my sister, and a drawing I created about something meaningful to me. When all three are done, the door turns green to show you’ve completed everything.It’s pretty straightforward, but that was intentional by the end. The goal became making something that actually works smoothly in VR instead of something overly complicated that breaks.

2) Development Process

The process was honestly a lot of trial and error and things not working the way I expected. I started with a much bigger idea where I wanted multiple types of puzzles, including word-based puzzles and color-flashing puzzles that would be more dynamic and interactive. But early on, I kept running into issues in Unity. Scripts wouldn’t behave how I expected, objects wouldn’t move properly, and sometimes things would just completely stop working. There were multiple points where Unity would crash or scenes wouldn’t load correctly, which made testing really slow and frustrating. Even basic mechanics took a while to stabilize. For example, I had issues with movement not working, triggers not activating, and doors not loading the next scene. I kept having to rewrite or simplify scripts just to get small things functioning.

At a certain point, it became clear that trying to build all the original ideas wasn’t realistic. The word puzzles and color-based mechanics were especially difficult to implement reliably in VR, so I decided to cut them. Instead, I focused on one core interaction and built around that. That’s when the project started actually coming together. Once I committed to a simpler system (snapping pieces into place), I could make progress without constantly breaking everything else. Even just getting the pieces to snap was difficult and I ended up having to find a work around for that. I want to continue to work on it to make it seem more realistic in the future.

3) Successes

The biggest success is that the game actually works from start to finish. The puzzles are functional, the snapping system feels intentional, and the door clearly shows when you’ve won. Getting the snapping system to work was a big moment because before that, interactions felt inconsistent and frustrating. Now it feels more controlled and understandable for the player. There are still issues, even in this version of the game, however I have definitely learnt a lot through this process. Also, even though I scaled back a lot, I still ended up with three distinct puzzles that feel cohesive instead of random.

4) Challenges

Most of the project was challenges, honestly. Unity crashing was a recurring issue, especially when testing scenes or changing scripts. That made it hard to tell whether something was broken because of my code or just the engine acting up. Debugging was also a huge challenge. Small issues like a missing collider, incorrect trigger settings, or a script not referencing the right object would completely stop my progress. When one thing worked something basic would break. A lot of time went into figuring out why something wasn’t happening instead of building new features.

Another big challenge was VR interaction itself. Things that seem simple (like picking up or placing objects) are much harder in VR because everything has to feel natural. And probably the biggest challenge was letting go of my original idea. I had a more creative and complex vision at the start, and it was frustrating to scale that down. But trying to force it would have meant not finishing anything.

5) What I Learned

I learned that scope matters a lot more than I thought. It’s easy to come up with ideas, but actually building them—especially in VR—is a completely different challenge.

I also got a lot better at debugging in Unity. I learned how important it is to check small things first and to isolate problems instead of changing everything at once.

Another big takeaway is that simple doesn’t mean bad. The final version is much simpler than what I planned, but it’s way more playable and clear. In VR especially, clarity is more important than complexity.

6) Possible Future Revisions

If I kept working on this, I’d try to reintroduce some of my original ideas but in a more creative way. For example, adding a simple version of a color-based puzzle or a basic word element, but only if it doesn’t break the core system. I’d also improve feedback more, like adding sounds or visual effects when pieces snap into place. Also a better reward when the game is finished. I'd also like to make the snapping objects more realistic by destroying the handheld object.