Initiating the v0.4 Overhaul

11/04/2025 : Initiated v0.4 Revamp

--------------

Intro:

Today I started revamping the 0.4 version of the game with a new vision for the game inspired by Neo Geo shmups and other arcade games. the past couple days I have been writing notes and trying to envision solutions to the game's current issues as I was getting increasingly confused and uncomfortable with the state of the game. After lots of mulling over, examining how other games felt to play, note-taking and testing, I came up with a list of potential causes for my frustration and possible ways to start addressing them. One of the main issues was that my game in its current state felt too chaotic and random in its gameplay as a result of lots of analogue values layering over the top of each other in the game's possibility space.

Too many possibilities:

In other words, the game had lots of variables that had a wide or continuous possible range of values. For example, the cursor could be aimed freely anywhere on the screen, and the 4 main gene values each had a possible value range from 0-9, just to name a couple. The 4 main genes have a total of 10,000 different possible combinations (different 'builds') on it's own, not to mention how that is then layered over the other gameplay dynamics at play at any given moment. This was done intentionally to start with exactly for the large number of possible combinations of states it presents.  I thought this would open up for lots of 'free' variation in gameplay and replayability, but I've come to feel now that it was a double-edged sword (hence the point of this revamp effort). It was creating too much 'mud' in the gameplay, moments where you just felt like saying "did that pickup really change anything significant?" or "Does trying to dodge even matter if everything moves randomly and I can't tell what's going on on screen?".

Getting out of the mud:

The gameplay was suffering from feeling too noisy and impossible to predict, especially in the enemy's AI and the projectiles. To try and alleviate the issues I decided to simplify all the systems in the game, starting with the enemies. I stripped out the new behaviour code I had started implementing that would change the enemy's movement and attack rate based on their genes. I reduced it back to a simple IDLE, MOVE, ATTACK, state machine instead and determined that enemies should only try to attack when they are idle. I also locked the enemy movement to the 8 cardinal directions instead of letting them move freely from point to point as another way of trying to reclaim order and predictability in the game.

Afterwards I moved onto the player and stripped out the analogue mouse controls - Now you have to use the arrow keys or face buttons on a controller to shoot, so it's locked to 4 directions now, for the same reason I restricted the enemy movement. I also reduced the range on the genes from 0-9 to a range of 0-4 (which reduces the number of possible combinations to around 650), making gains in each of your stats feel more noticeable and impactful than before. I might change it back in the future, but for now it makes it a little more manageable to work with and design around.

Meaningful variation:

My current issues aren't with variation itself, but the fact that the variation was unstructured. The variation wasn't meaningful variation, but just the creation of more possibilities, introducing more noise into the system. To re-introduce variation into the game in a more controlled way, I've started implementing different enemy types, such as a turret enemy that stays put and shoots in a big spiral, an orbiting enemy that orbits around the player to avoid attacks and take pot-shots, and a 'hyper' enemy that will rush around and shoot unpredictably but drop lots of currency on death.

Conclusion:

We'll see how the game feels again in a few days, but so far I think the changes are starting to bring the gameplay back under the control of design and it's starting to feel fun to play again.