Project Postmortem

What did we change from the original plan?

We initially planned for the Lumintrail's main gameplay loop to focus on staying on a winding path, but we decided to pivot due to the challenges of creating curved paths in Unity. While we eventually managed to implement curved paths for levels 2 and 3, it proved to be quite difficult yet educational. As a result, we leaned into designing a more challenging obstacle-dodging game. Additionally, we planned for the player to ride a skateboard, but we later changed it to a snowboard. With the skateboard, we wanted to put more focus on the movement by allowing the player to drift and perform tricks with an animated skateboard. However, due to time constraints and assignment deadlines, we opted for a snowboard with simple and smooth sliding mechanics.

What went right with our final project?

One of the biggest successes with our final project was how quickly we got a bare-bones version of the game up and running. We emphasized the importance of focusing on core gameplay mechanics and the gameplay loop before adding visual polish or assets, and it served as a clear representation of our vision for the final game. Even in its early stages, the prototype effectively captured the core fantasy we wanted to deliver.

What went wrong with our final project?

One of the main challenges with our final project was our limited experience with Unity. Initially, we assumed that creating curved paths would be straightforward, given how common they are in games. However, we quickly realized this wasn’t the case. For our prototype, we opted to use a simple slope with randomly generated obstacles, which turned out to be surprisingly fun. The randomness meant the wisp chose a different path each time, keeping the gameplay fresh.

Later, we attempted to incorporate curves using the spline package, but the paths it generated had invisible edges that caused the player to get stuck. After researching on YouTube, we discovered the Easy Roads 3D asset, which resolved the path creation issue. However, integrating it with the NavMesh for the wisp’s AI pathfinding introduced additional complications. While we eventually got everything to work together, the wisp still occasionally got stuck on corners, and despite our efforts, we were only able to reduce—not completely eliminate—this issue.

We also encountered problems transitioning from cutscenes to gameplay, but Giorgio and Nate managed to resolve these challenges successfully, ensuring smooth transitions in the final version.

What would we do differently next time?

If we were to approach this project again, we would likely create our own curved road assets using external software like Blender or use the spline system solely for the wisp’s navigation to prevent it from running into obstacles. However, we also encountered challenges with timing the wisp’s navigation to start only after the cutscene ended, which would require additional refinement.

Another change we would make is starting with Unity’s URP from the beginning. Transitioning from the Built-In Render Pipeline to URP later in development caused issues with shaders and materials, which could have been avoided with a more consistent pipeline choice from the outset.

If we could add one more week of development, what would we add?

If we had an additional week of development, we would focus on adding more variety to the game. While we currently have three slope levels, we envisioned introducing new environments, such as a cave level with stalagmites, stalactites, and maybe some jumps, to expand the gameplay into a more full-fledged level-based experience. Another idea we would explore is implementing an endless mode, as most of the code for it is already in place. This mode could be a special feature within the game, similar to Google’s Dino Run, where the environment changes periodically to keep things dynamic and engaging.

A simpler yet impactful change would be addressing the issue of the wisp occasionally hitting obstacles and getting stuck. While this is largely due to limitations in the NavMesh’s pathfinding, it would be our responsibility to find a workaround or fix. Additionally, we’d like to incorporate coins and upgrades, which would complement the gameplay well. Coins could unlock cosmetic options, while stars collected during levels could be used to enhance the wisp’s light intensity and range, adding a layer of progression and customization to the game.

Team Member Contributions:

Everybody equally contributed to brainstorming and gameplay decisions such as how the game is played.

• David Guerrero-Pantoja
  • Player Movement, Lighting mechanics, Level design, Saving loading functionality, displayed star UI on win and level select, path creation, Debugging, trailer video, collectible collecting
• Joshua Hall
  • Updated aesthetics for menus, Navmesh and Wisp AI, Level Design, Camera, Damage to player, particle effects, snowboard trail, Debugging, trailer video, death animation, general juice stuff, itch.io webGl build, options menu
• Nathan Shturm
  • Player, Trees, Collectable, Crystals, and Snowboard assets, skeleton for intro cutscene, countdown animation, materials and shaders for objects, obstacle spawning, death barrier
• Giorgio Vallesi
  • Base menu scenes, menu canvases, all sounds, finished intro cutscene and countdown, pause menu