Inconsistency is Beautiful: In Defense of Fighting Game Jank

September 7th, 2023

This is a repost of an article from my cohost, posted on august 23rd, 2023. People seemed to like it a lot though, so I'm reposting it on my blog.

Gonna babble for a bit and hope this is coherent:

I was weirdly saddened today reading Strive's patch notes. A removal of the character weight system. A younger version of me would be SO RELIEVED by this. "Oh god I don't have to memorize a million different combos"! Yet now, an older me, is oddly sad?

Now, I'm not gonna hate on or argue about Strive, or any other game. Plenty of games I like have equal character weights and consistent hurt boxes. I'd rather game designers do what they want to do, rather than pander to me.

(Granted, I do wish more people were pandering to me, but that's a me problem.)

... Instead I want to be more positive about the stuff. So much of this conversation gets caught up in arguments about gatekeeping and "git gud" "Baby Game" BS but not a lot of people really go into why they might like some of these arcane systems.

A nice and polite twitter follower, immediately after I tweeted my disappointment, asked...

Why would you want combos to fail randomly per character performed on

... which like lol, when you put it like that, it sounds super silly. But it's that framing -- a framing I've seen many times. I remember being on a forum... very appropriately, it was David Sirlin's forum (thank god you can't name search on cohost)(edit: uh oh). Being the Sirlin forums, you expect a... certain type of person and player. Very big anti-execution crowd and I was like the only real execution defender (at least who was a semi respected member of the community and not a random SRK troll). I remember one exchange talking about GG combos and the comment "Well what's fun about just doing the same rote thing over and over again?"

"Well you're not? Like I'm adjusting my combos as we go, depending on how high they are and stuff"

"I don't believe you."

Now, this is mid 2000s. I don't think anyone now would deny that's a thing that players do... but I think it still highlights a way a lot of people still feel. Combos as this discrete thing, these bits of work you get through to get to the Real Game (that forum LOVED talking about the "Real Game"). You learn your combos, so you get to play brain chess.

But instead the whole thing is very fluid, especially in a system rich game like the older Guilty Gears. You never stop learning, and that combo you learn isn't a discrete unit. It's a lot of different smaller parts and that perfect hit you need to do your idealized BnB is actually kinda hard to land. You need to learn how to put these things together in different ways. Combos are less raw memorization, and more a matter of a little memorization, but a lot of developed intuition.

This is no surprise to anyone whose played a lot of really nutty fighting games. But the important thing is more the mentality of "Combos are a thing that you need to have, and you fucked up if you weren't optimal" vs looking then as an extra and not taking them for granted.

"... Wait, can I convert to this route off this hit?"

Often in games with open ended combo, you'll get a hit and you won't actually know what you can get off it. I recognized the situation 3 hits in.. what's the gravity scaling like? What's their character weight? This route doesn't work on her hitboxes usually, but I think it might because of the weird height I hit at??

From there you gotta bet on yourself. Take the easy knockdown? Try to extend to a damaging route? What are the stakes of the match? How much life do you have? Is it worth maybe eating shit just to find out? Those sorts of situational, high speed valuation processes, for some people like me, are extremely fun and with games like +R or Rev2, I'm still, after thousands of hours, guessing and developing my intuition. Every matchup has new things to teach me not only in neutral, but on what to do when I even hit someone.

I don't like character weight because I like dropping my BnBs, or because I want to make the game harder for new players, but because they always keep me on my toes and give me great moments where I am rewarded for my intuition. I like it because I can do cooler combos.

... What if input buffers made games harder?

I was playing one day with Lofo, a really incredible +R Dizzy player and a former (lol, recovered?) Sirlin forum poster who ended up a huge execution lover. One day we're talking about Rev2 vs +R and hit me with something that has been in my head for like 2 years. Something to the extent of...

"Yeah, I don't like Rev2's input buffer. I feel like it makes the game harder, because everything is more consistent... I... don't think I like input buffers?"

Which to me at the moment felt like an insane position. Like there was a lot of simplifications made to fighting games I didn't like, but that one seemed like a clear win. That just makes games better, right?

But Lofo kept talking, about things that are borderline impossible in +R that would be consistent in Xrd and how one of the things that keeps +R reasonable is that everyone drops stuff all the time. Not just in combos, but in pressure. There is always wiggle room... and then talking about mashing to tech.

Mashing to tech feels like a vestigial part of Xrd. It doesn't bother me much (I come from X2), but if you're trying to tech and there's a gap, you're gonna get it. +R, much less so. It's almost an analog skill check between you and your opponent. Your ability to mash, vs their timing during the hardest parts of their combo. Defender can piano, so there is a bit of an advantage

Then that got me thinking about ST. "It's fucked up that you need to do a 1f reversal to beat tick throws in that game."

... But you don't. You need to have better timing than your opponent to beat tick throws. Can they time to 1f input? If you're playing someone great, probably, but when you watch mid level play, most DPed tick throw attempts aren't usually reversals. That analog sense of timing is part of the game's skill expression.

This goes into why people didn't care about exact frame data back in the day or players playing "by feel". A move being +1 really didn't matter unless both of you have sick timing. We HAD the frame data. We had Yoga Book Hyper for ST. It did help. But it's influence was different because the play conditions are were different.

In modern games, a +1 situation is often pretty rigidly defined. We have buffers. Our responses will come out on he fastest frame. If my opponent is slow and my suboptimal option keeps winning, people will call that fake... because it is. The expectation is that verse most players, even low ranked players, people will get their moves out as soon as possible. Meanwhile in older games, you can't take that as a certainty even with the best players. They'll hit a lot of frame perfect inputs, but not all of them. Finding where your opponent is being sloppy helps a ton. No one is clean all the time even in modern games, but it's so SO much harder in old games.

I even think a lot about setplay characters. In older games 'perfect knockdown into oki that grants an auto timed safe jump' is actually super hard (or really lucky happenstance). Heck, this is also where GG's variable wake up timing stuff also comes in. You could do it, but it would be so hard that it can never be the expectation. Now safe jumps are so easy once labbbed that if you whiff a normal before doing your oki people will just assume it's a safe jump even if it isn't. You get stronger setplay because frame perfect repeatability, while not at all trivial, is extremely practical.

Buffers help turn is into robots and, depending on your taste, that can be a good or bad thing.


One thing that I've also thought a lot about is... new players seem to have an easier time getting into +R than Rev2? Part of this might also be the lobby system and speed to matches, but part of it is, in Rev2, even a mid level player can be pretty scarily consistent, but +R... Welcome to the scramble zone, lol. And like granted you can run into cryptids with 10,000+ hours of play time who will Burst Safe Sidewinder Loop you into the negaverse, but even THEY fuck up or get wilded out by weird interactions. And I say this maybe liking Rev2 more than +R.

In a weird way, making games easier, also makes them harder, because you make them more consistent for everyone... and when everything is more consistent, the game is more rigid and unyielding. You're not making an old experience accessible to new people, you're making something new, with it's own pros and cons.

Again, this isn't a judgment zone. I'm okay with Strive. I'm actively loving SF6. But a rigid games forces players to play it how it was intended. This can help new players learn a lot faster. Hell, such design has lead to games that have even taught me lots of stuff! I don't hate these games.

... But I miss that looseness. I miss how you can have a combo so hard that only like 2 people can do it reliably and just this really hazy, unclear idea of what's even possible. Infinite weird, crufty interactions between interactions. Feeling like I wasn't just playing my opponent, but exploring a rich, emergent design space.

Fighting games as a genre increasingly feel like they're (metaphorically) moving from "analog" to "digital".. and like most of those changes, there are usually more advantages and disadvantages, but, even with the new advantages... there are always gonna be people who miss how the old analog models used to feel.