paosditygasfdio;yhtglkadfhtl;q3h4poghdfc
this post doesn't really have much useful information on it but i want to log the creation of this game for prosperity, please bear with me!
hypothesis
i think fighting games rock, but it's 1v1 modes haven't diverged that far from the main formula discovered in street fighter 2
raph koster made a really interesting point that fighters are essentially rock/paper/scissors played on an axis
https://youtu.be/zyVTxGpEO30?t=1430
playing fighters since forever, this rock/paper/scissors analogy is totally true. pretty much every option has a well-defined way to beat it — you can anti-air jumps, you can jab slow overheads, etc. all that matters is having the gamesense to know what's coming out at what time and being proactive, or being fast enough to be reactive. it's an environment where you don't have time to sit and meditate on your response to an opponent's actions, so your body is constantly bordering conscious and subconscious operation which means you - human being - will play out predefined patterns, and your opponent can adjust their behaviour accordingly
there's been many games that diverge from this formula and are incredible, like divekick, lethal league, and recently yomi hustle
yomi hustle is godlike and a successful experiment in removing that immediate reactive time pressure, deconstructing the genre and reconstructing it into a turn-based game — it still has that insane fighting game spirit of prediction, learning what moves you can string together to optimize your output off of a correct read, and so on. it cuts the need for time-based execution and keeps a lot of fighting game ideas intact like combos and mindgames
https://ivysly.itch.io/yomi-hustle
divekick is also a nuts deconstruction of fighting games, you get literally 2 moves. its a distillation of the last few tense seconds of a match where both players are at 1hp
it essentially becomes completely reaction, mindgames and each character's unique movement. this game made it super far, having regular tournaments of people who were doing stuff that looked insane, so this was a super successful experiment too
so is it possible to make a new deconstruction that removes both intense timing-based execution and combos, but tries to keep realtime mindgames, and still be fun?
the game
i made a fighter that is quite literally rock/paper/scissors
you have 3 moves: punch, parry, and grab
punch beats grab, grab beats parry, parry beats punch. this is the RPS, and the axis is the 2 dimensions of movement
jumping in this game is a huge commitment. you can't change your direction during a jump, but you can still attack as a counter to a ground attack
one thing i wanted to capture is an idea that comes from smash where characters can get a KO very very early. i wanted that same vibe, but sticking to the ground - so i wanted to go for a sumo-style game where you KO your opponent if you push them into a wall. rather than decrease health or increase knockback percent, all attacks push your opponent back closer to this wall, where they can take risk to get back to safety, or stay in this tense disadvantagious spot
playtesting the finished game!!!
the majority of the work for this game was the shift to 3D (!!! which is hard when you're working in a framework built for 2D lol), and a refactor of the rollback netcode container. the rollback rewrite was to fix an off-by-one issue. turns out that happened due to some internal game state not being serialized correctly but i reduced the size of the rollback container by 50!! so big win regardless
big thanks to geni and ceroribi for playtesting this with me & giving feedback!
i was surprised that the game already felt pretty fun! on paper it feels like it should constantly be a game of random chance, but it's very clear that you can get good at this game and be better than the opponent, which is great. the slow-paced deliberate (and clunky) feeling is great. it rewards patience and mindgames over quick reflexes, execution and button-mashing. if you press a random button, you will very easily get punished for it
the sumo-style works very well, you can safely control centre stage or hover around a wall to get a quick kill, while being at a huge risk yourself. you get used to the clunkiness, but it is missing some simple game-feely stuff, like vfx, sound, and animations that aren't terrible(!!!). i feel like if this game was given a budget, it could be a fun small argument-settler
feedback
the game has traditional blocking where if you hold "back", you don't get hurt. this sucked, we already have a defensive option of parrying. blocking is too little risk. we can remove this
crouching punch isn't worth anything, the focus definitely wasn't on mixing up high/lows. instead we can turn this crouch punch into an uppercut anti-air
as a counter to the uppercut anti-air (and also an extension to the sorta limited options you have while you're airborne), we should be able to both parry and grab while in the air
you should be able to dash while in the air
charged attacks! if you anticipate your opponent reacting with a parry to your punch, you can charge it and wait for the parry to end to release it. if ur opponent anticipates your charge, they can grab you or wait out the parry. like smash!
im pretty happy with this project, which is nice, since partway through i was certain this idea didn't work at all and was pretty bummed about my efforts. im still unsure, but friends playing it and having a nice time was really reassuring, so im trying to power through that feeling a little more
ill let it sit for a while before i refine it and apply these new ideas. thanks for coming!!