I finally decided on my game's title and basic premise - Armed Decobot (as in decontamination robot). A game about a futuristic street cleaner who's trying to recover her stolen van using a cleaning mech/ride armor repurposed for combat. I might tweak this a little but otherwise I'm committing.
Now that it's settled, I can actually work on fleshing out the visuals & story of my game and replacing the temp assets with something that looks a lot nicer.
This happened a short while after I figured out the game's overall direction after almost a year of aimless flailing. Initially it was meant to be more focused shmup-like game. Then I awakened my inner technocrat and decided to try to add bloat to the game for the sake of some imaginary "beat 'em up normie" audience.
After snapping out of it, I've almost entirely ditched the whole "bigness" crap, scaled back the amount of planned content and focused on fleshing out the core interactions. I've also been increasingly focusing on the beat 'em up aspects of the game. The shooting is there, but it's niche and primarily used to "open enemies up" rather than being an alternative to melee. The combat's starting to come together!
Now that I got rid of the stupid self-created problems, I'm now left with a lot of far more difficult problems I have to figure out solutions to, ones which are inherent to the type of game I'm making, or games in general.
The biggest, most persistent problem, and one I want to focus on in this devlog is : The player can't turn around.
This might seem like a relatively small thing, but it's not. I'm all about trying to give the enemies a game plan, I want enemies to constantly be trying to do something the player finds undesirable. However, this process should always be two sided and granular - the player should always be able to win back advantage in a fun way, without immediately getting checkmated or doing awkward crap to recover. The goal is to ramp up tension/excitement & give every action meaning, not just to make things difficult or punish the player, after all.
Usually, in more typical "genre" style games, strong game plans are inherited. They have strong sets of fundamentals which have been built from years or decades of iteration.
In shmups, you are trying to push up and quick kill enemies while they want to push you down and corner you. The competiton between player and enemies is always happening and the entirety of a player's mechanics are being used to engage in it.
In beat 'em ups, your goal is to keep enemies in front of you while their goal is to get behind you. Because you can turn around, you also get to use every mechanic to participate in the tug of war, even if you're on the losing side.
But now what happens when I combine the "only look ahead" nature of shmups and the enemy game plans of beat 'em ups, breaking away from each genre somewhat? It creates an uncontested space behind the player. If enemies slip in behind you, you have to rely on niche properties of your moves to get them out - you're now playing a fraction of the game you were playing before they got in. Even if you do it and scoop them out, it just doesn't feel good or cool in any way. There's a reason why something like Breakout kills you when the object gets behind you - it's better to reset the game than letting the player go through the frustrating nightmare of scooping the ball out somehow.
As a result, the bmuppyness of the enemies' game plan had to be minimized - they had to restrain themselves when getting behind you. If I give the player tools to do it effectively, it ends up being very situational and unpleasant, simon says type of shit. And of course, because the game is heavily state-based, I can't go wild with bullets because they will constantly interrupt your actions. So in a way I am locking myself out of the natural ways to create enemy game plans that come pre-packaged with both genres.
In exchange, I'm trying to emphasize angles. The enemy's "win condition" is attacking you from an uncomfortable diagonal angle, or from the side. This way, most of your attacks could still hit, and you could slightly reposition to dodge the attack & retaliate, but it wouldn't be trivial.
While this worked, I was still feeling bummed out by how over-reliant I was on angles or ridiculous hitboxes for any sorta challenge, and how static it all felt. That's when the Footsies Skeleton came to the rescue. It really highlighted to my dumbass brain that there's a lot more juice to squeeze out even from "vertical" interactions with enemies - I just have to focus more on creating interesting movement patterns for enemies.
This encouraged me to do more proper research on beat 'em up enemy AI, and learning how their movement patterns worked inspired me a lot and gave me a lot of ideas on how to create interesting movement dynamics & game plans without relying on enemies getting behind the player. Along with how to emphasize the diagonals. You can see what I did here - a repulsive square moves around based on the player's inputs and gives enemies a singal that they use to decide if they want to move away, or move towards the player. Later, I even added another static repulsive box behind the player, to avoid them becoming too much of a pain to deal with.
So now it's just a matter of creating a lot of fun varied enemies using this concept!
Some notes for myself, describing other issues I wanna talk about later :
- Dodge-style abilities are interesting in multi-enemy encounters, but mess up a lot of 1v1 dynamics.
- "Advanced" enemy moves/techniques like blocking/rolling are very disruptive to the crowd control based gameplay.
- Controlling where the enemy falls is too chaotic, unlike beat 'em ups.
- Looping attack sequences pop up constantly and are hard to get away from.
- It's hard to "contain" move ranges - ranged weapons are insanely dangerous.
- Motion Inputs & shmup movement don't mix.
