Project Post Mortem (Month 1 Vertical Slice )
Introduction: Vira Bloom is a PC topdown semi-roguelike taking place in a microscopic biological environment, of blood vessels and cells. The player is shrunk down into a nano capsule that fires projectiles and can improve their capsule’s features based on items collected in the level that can enhance and help the player get past the main objectives.
Link To Gameplay Video: https://youtu.be/AL_KyKBZr60
What Went Right:
Character Movement customizing- I wanted the topdown to have a scifi, but similar to roguelike feel, and overall I’m pleased with how the movement, camera, and feel of the character moving around the world went. It absolutely took some testing and adjusting to get camera position, and controls right. I wanted it to be simple so click and hold was my desired go-to. It was loosely based on how the UE top down was implemented, except the player isn’t automatically going when you click. Giving a more intentional movement for when the game gets more difficult. I figured this would be the best way to boilerplate some basic player movement if I wanted to also port to mobile, so having it simple with just mouse and clicking was important.
Theme Appearance. I envisioned a scifi, with elements of biology, science and physiology. I used general information about cells, blood vessels, and was able to make a texture that I felt resembled the inside of a blood vessel or intestine very well. Not necessarily the human body, but an animal body microscopic environment. I used a landscape to carve out channels, corridors, and areas of refuge.
Core Mechanic Shooting- The general click point and shoot was implemented to how I wanted it, not too far off from common PC roguelikes i've played on steam like “Ream of The Mad God”. Granted the BASIC feature of how shooting worked is the ONLY thing that went right, more on that in the latter sections.
Core Mechanic Collecting - A simple approach to collect something was simple, easy to use, and communicated well with sounds, UI features, and how they were implemented, especially the “nutrient” for power the blood cells would drop after destruction.
Progression, Game flow. - To get to the main cell and destroy the virus, you needed a sodium ion to get past the “gate” to get inside the cell was easy to code and worked as intended. The level has a start, middle, end, with opportunities of exploration, collection, and optional activities that are scalable and adjustable for future levels.
What Went Wrong:
Projectile Bad Boiler Plate - How I implemented the projectiles was really bad for scalability, modularity and versatility in the game. I was very hardcoded with how they spawned, what damage they did, and it made it very difficult to change it later in the development, which really made poor implementation of the “buffs” feature via the collectables mechanic. Had I used structs/DataTables, or Tags, to better organize from a projectile parent class and what data was being managed, I think it would have made it much easier to make adjustments in the time given. Each projectile had to be its own blueprint and instead I should have had a projectile parent, with components that could have been changed based on conditions the player met in the level.
Level Communication - Per player feedback, the more experienced gamers didn’t seem to understand the point, there weren't early rewards, and the settings for damage were rather tedious, and communication between enemy and red blood cell differences really hurt the intended theme. A stronger lean into the environment, adding more subtle differences to highlight the different zones, pathways, and even a minimap would have made navigation and level communication more helpful.
Theme and Aesthetic Improvement - While all players generally understood without prior knowledge the game was about being in some microscopic biological environment , there could have been more graphical things made in the “refuge” parts of the level to communicate where the player was and what was happening around them.
Enemy Variation, Aesthetic, Punishment - The enemy were these dark red triangular things that weren’t really clear on what they actually were. The boss was just a larger version of the basic default enemy. The enemies did not chase, or attack back. Future updates should include varying enemy actions, navmesh, behaviors and punishing features for the player to negotiate.
Initial intended mechanic - scope creep. This can also be attributed to bad boilerplate code and planning. Initially I wanted several different things to collect, and different rewards/features for the player to access based on what they collected, but it ended up becoming hardcore spagetticode fast, and performance was also taking a hit, so I had to delete a lot of cool planned features from the beginning. Really starting out with a simple principled mechanic and building from there instead of making conflicting features and things all at once and assuming I could make them work in the end was really really bad, keeping to the scope of a single level, and focusing more on the IPM model of mechanics would have not only simplified this but probably given more interest and purpose for my more hardcore testers that didn’t find the game very interesting.
. Back to main