Marvel vs. Capcom 2 is the progenitor of a particular fighting game subgenre. I don’t know that there’s a single name that we’ve all agreed on, but I’m going to stick with calling these versus games (I don’t like terms like “tag games” since King of Fighters, Tekken Tag, etc. are their own separate thing). As a kid, I had Marvel vs. Capcom 2 for my Dreamcast, but I had no idea competitive fighting games were a thing, so I just had a great time playing arcade mode for hours on end. My introduction to playing versus games “correctly” was its sequel, Marvel vs. Capcom 3, back in 2011.

Ever since Marvel 2 established the core of this subgenre and became a phenomenon, more and more developers have been throwing their hat in the ring, trying to capture that versus game essence while putting their own twist on it. My view of the family tree is something like this:

• Generation 1

  • Marvel vs. Capcom 2
    Note: I know Marvel vs. Capcom 2 was built on the foundation laid in Marvel vs. Capcom, X-Men vs. Street Fighter, etc. I still consider Marvel 2 to be when things really crystalized here.

• Generation 2

  • Marvel vs. Capcom 3 (2011)

  • Skullgirls (2012)

• Generation 3

  • Marvel vs. Capcom Infinite (2017)

  • Dragon Ball FighterZ (2018)

  • BlazBlue: Cross Tag Battle (2018)

  • Power Rangers: Battle for the Grid (2019)

• Generation 4

  • Hunter x Hunter: Nen x Impact (2025)

  • 2XKO (2026)

  • Marvel Tōkon: Fighting Souls (2026)

  • Invincible VS (2026)

  • Scramble Heart City (TBD)

I think both Gen 2 games were really vying for the recognition of being the “real sequel” to Marvel 2 (whatever that means). While Skullgirls obviously didn’t have the Marvel license, they were clearly patterning many of its mechanics directly after Marvel 2. Ultimately, both games were successful in their own way, proving Marvel 2 was not just a one-off success, and this paved the way for the explosion of versus games we’ve seen in subsequent years.

As I mentioned earlier, Marvel 3 was my introduction to this type of game. I absolutely loved it, and I’ve been chasing the dragon ever since. I wrote a post not too long ago about how I still don’t believe there’s been a game that quite scratches that same itch.This isn’t a value judgement about how good or bad any of these other games are; I’m just saying that having assists or DHCs or whatever else will not necessarily mean your game actual feels anything like Marvel to play. This is fine! (though I selfishly want more Marvel 3 🙃)

We’re now in the midst of the Gen 4 releases, each with their own interpretation of the subgenre.

Enter Scramble Heart City

Scramble Heart City is a 3v3 versus game you can play in a web browser from the team behind Tough Love Arena. They’ve been doing open betas one weekend every month, and the December 2025 beta is the first one where I’ve had time to sit down with the game. Going into the game, my two questions were:

1. Where does this fit in that whole versus game genealogy I talked about earlier?

2. Comparisons to other versus games be damned—how much do I like this game on its own merits?

Ultimately, question two is the more important one. Scramble Heart City evokes Marvel mechanically in a bunch of different ways. You have a three character team with the player choosing one of three assists for each character. You have DHCs. You have pushblocks. You activate pushblock with the same button combination you use to dash. You can wavedash by repeatedly dashing and then cancelling the dash with a crouch. You can jump cancel your dash. Characters have varied movement options with different combinations and versions of double jumps, air dashes, and flight. There are install supers. It’s a six button game with light attack, medium attack, heavy attack, special, assist 1, and assist 2. There is a ton of Marvel DNA here. This game doesn’t venture into the active tag territory that MvC:I, BBTAG, and 2XKO have explored and sticks much closer to the Marvel 2 roots of the genre.

What I Like

Scramble Heart City uses one button for special moves rather than traditional motion inputs. More fighting games have moved in this direction in recent years, and I have mixed feelings about it. I made a video about how I mostly think one button specials are a good thing because they make a game play well across as many input devices as possible. That being said, I think they introduce certain problems even as they solve others. There’s the old adage of “Good, fast, cheap. Choose two.” The fighting game controls version of this is:

1. Expansive movesets

2. A reasonable number of buttons

3. Simple controls

You can really only have two of these. Traditionally, fighting games used motion inputs so that we could have both expansive movements and a reasonable number of buttons. With just four attack buttons, Dante in Marvel 3 has something like 50 unique moves. If you want to keep expansive movesets and also have simple controls, you need a million buttons. I personally dislike this approach. It’s one of the main reasons I bounced off 2XKO as hard as I did.

If you want simple controls and a reasonable number of buttons, you need to do more with less and make a game that’s fun and deep without characters needing dozens and dozens of moves. Scramble Heart City takes this approach, and I think it works really well here. The characters feel distinct and mechanically unique from one another.

Wavedashing, air dashing, and double jumping around also feel great. If a fighting game feels bad to move around it, I probably won’t stick around for long. Moving around in Scramble Heart City is a joy.

The game has some universal combo structures, so with most (all?) of the characters, you can get up and running with something like universal ground chain → air chain → air super. These aren’t optimal combos, but this lets you try out different teams pretty easily, and versus games are built on this kind of experimentation.

When the beta started up, I hopped into training mode thinking I’d spend maybe a half hour getting a feel for the game to see if it was anything I’d even be interested in. I think I accidentally spent something like four hours in training mode just noodling around, something I really haven’t done since Marvel 3. This game feels good to play, full stop.

What I Don’t Like

Since the genesis of the subgenre, versus games have had some degenerate shit. To some extent, this is a feature and not a bug. If there’s enough competing degenerate mechanics and tactics, it can kind of all even out and just work somehow. The power levels of the characters in versus games are deliberately orders of magnitude beyond what you’d see in a Street Fighter game, and these absurd power levels are part of the appeal.

That being said, degen stuff can also sometimes break down, flattening the game down to a single dominant strategy (making things boring). As much as I love Marvel 3, I enjoyed it despite things like TAC infinites, Zero 300%ing a team, or dark Vergil melting an entire team in 10 seconds by himself.

While Scramble Heart City has pushblock (and a more powerful version of pushblock you can do by spending a meter) as a defensive mechanic, right now it feels like it can become a one player game once someone who knows what they’re doing lands a hit. Once knocked down (best I can tell) you have no control over your wakeup timing and you have no tech roll options, so if you survive a combo, you need to guess some coin flips to earn the right to play again. Obviously oki is not unique to versus games—my issue here is more with the risk/reward for the attacker and how many coin flips the defender needs to correctly guess in a row in order to play again. Right now it feels like there might be too little risk, too much reward, and too many coin flips.

After a character death, your incoming character is put into this same mix-up situation, so the game can easily become a single neutral interaction into looped oki into ggs.

The Rest

I have some other random thoughts, so here they are in no particular order:

Character Count Could be a Problem

There are currently five playable characters, and there are three additional slots blocked off on the character select screen. I don’t need or expect 40+ characters, but eight characters feels very sparse. I hope they’re eventually able to get to double digits at least.

Does Team Comp Matter?

One of the coolest things about versus games (when done well) is how the unit of play is your team, rather than any one individual character. Quoting from an older post I wrote on this same topic, “I think the synergies that make your characters cohere together as a team are what makes versus games truly special. If you can pick any random three characters and it works alright, it probably means synergies just don’t matter all that much.”

Using a game I know well as an example, Marvel 3 has a bunch of team comp elements that matter a ton. Having a first character with a strong DHC out (reversals like Spencer’s bionic arm super are great for this) and then a second character with a strong DHC in (safe supers like Storm’s hail storm or installs like Dante’s devil trigger are good here) can form the backbone of a team. Team supers like the Doom/Amaterasu follow-up my lead can be cool team-specific meter dumps. Assists are arguably the most important here—playing Magneto with Doom missiles assist feels meaningfully different from playing Magneto with Sentinel drones assist feels meaningfully different from playing Magneto with Rocket Raccoon log trap assist.

I’ve spent so little time with Scramble Heart City that the jury is still out on how well they’ll pull this off, but I’m hopeful that they will.

The Slippery Slope Problem

Historically, fighting games don’t have issues with slippery slopes—they’re not over until they’re over. Like in Street Fighter, a Ryu at 50% health has all of his moves and combos and is just as dangerous as Ryu at full health. This is not true in versus games—as you lose characters, you lose tools. Ever since Marvel 2 had players quitting back to character select when they were down to solo Captain Commando, versus games have been trying different solutions to this problem, with varying degrees of success. Marvel 3 had x-factor. DBFZ has sparking blast/limit break. Currently, Scramble Heart City is sticking closely to the Marvel 2 mold of “if you’re down to your last character, tough shit.” Solo characters can DHC into themselves (so you at least get solid damage output if you find a hit with your last character), but I don’t think that’ll do much to mitigate the slippery slope. For better or for worse, when you’re down too your last character against your opponent’s full team in Scramble Heart City, there’s a good chance you’re cooked.

Infinite Prevention System

Some versus games use a system called hit stun deterioration (HSD) to limit combo length and (hopefully) prevent infinite combos. HSD means that lots of stuff combos together pretty easily early on in the combo, and then as the combo goes on, moves do less and less hit stun, and eventually even your hardest hitting moves don’t have enough hit stun for another attack to connect, so the combo must eventually end.

I like HSD. HSD (along with damage scaling) tends to result in combos having diminishing returns. A really basic combo might do like 60% of the damage of an optimal combo. With a bit of practice, you can do a combo that’s not too much more difficult that gets you 70% damage. Players with great execution are still rewarded, but they’re not doing double or triple the damage of someone using fairly basic combos. It’s also pretty intuitive (to me, anyway). As the combo goes on, you start omitting light attacks once they stop working. Then you omit medium attacks. Then you find a way to get a knockdown or combo into a super before HSD puts a hard stop on things. It’s simple and makes sense.

Scramble Heart City limits combo length via an infinite prevention system (IPS). Pioneered by Skullgirls, IPS was a system meant to prevent infinites by detecting when you do the same move sequence in a combo more than once. Scramble Heart City’s version of this has a UI element that appears once a combo gets going that fills as you continue the combo. It fills faster depending on the type of moves you do and how many times they’ve already been used in the combo. If the the bar fills all the way up, the character pops out of the combo and gets a bar of super meter (as the attacker, you really want to end the combo right before IPS ends it for you). I don’t love IPS as a concept. When I’m trying to lab out a new combo, it just feels like I’m grasping at a math equation I can’t see, trying to read the designer’s mind and guess their intent.

Conclusion

I honestly don’t know where I’ll land on this game. It’s so early in development that it could become a pretty different game six months or a year from now, and many of my criticisms might be rendered moot. At very least, I’m having fun with it, and I’m excited to see where it ends up.

Of all the fighting games I’ve put any significant amount of time into, Vampire Savior is my personal favorite. The Vsav community is amazing, so when I was just starting out, there was no shortage of folks who were happy to answer any questions I had. I later realized that there were things I didn’t even think to ask about (or things that I should’ve asked about sooner). It’s hard to escape the Dunning-Kruger Effect!

This post is aimed at fighting game players coming to Vsav from other games. Fighting game knowledge is very transferable, but while existing fighting game players can pick up a new game much more quickly than someone brand new to the genre, they also bring a lot of baggage and expectations from the other games they’ve played, which can sometimes hinder the learning process in a new game. The following is a non-exhaustive list of things I wish I had known when I was first starting out in Vsav.

A Few Things I Wish I Had Known Sooner

1. (Almost) Everything is Plus on Block

In many modern 2D fighting games, it’s a relatively safe assumption that any given grounded normal is probably at least a bit negative on block. In Guilty Gear Strive, Sol’s c.S is +3 on block, and his f.S is +2 on block. All of his other grounded normals leave you at some kind of frame disadvantage. Street Fighter 6 is similar, where unless you’re using driving rush or mashing jabs, a character will have one or maybe two grounded normals that are plus on block.

In Strive or SF6, you can assume grounded moves are minus on block and memorize the small handful of exceptions. Vsav, like many older fighting games, is the exact opposite. For example, here’s a list of Felicia’s grounded normals and their frame advantage on block:

Yes, her sweep is +3 on block. How does this not break the game? First, like many older 2D fighting games, pushback is pretty significant, so even with this kind of frame advantage, the attacker may be far enough away that they’re less immediately threatening than you might expect. Second, and more importantly, Vsav gives you strong defensive options while in blockstun, which brings us to point number two.

2. Start Your Pushblock Practice!

One of the core defensive mechanics of Vsav is advancing guard, which people most commonly refer to as pushblock. The name makes intuitive sense—a pushblock is the ability to push the attacker away when you’re blocking their attack. If you’re watching a Vsav tournament, you’ll hear commentators interchangeably use the terms pushblock, advancing guard (which they might shorten to “AG”), and tech hit (the in-game voice says/displays “tech hit” upon successful pushblock).

A pushblock or two will generally get you some breathing room and get you back to neutral—it’s incredibly strong! Remember that chart of Felicia frame data we just looked at a second ago? Pushblock is one of the reasons it’s a lot less scary than it might seem at first glance. It’s also one of the hurdles to clear to become an intermediate player in the game. I call it a hurdle because the input for pushblock is kind of weird, so it’s probably going to take you some time to figure out a method that works for you and get decent at it.

You pushblock by mashing buttons while blocking. Every additional button you press increases the odds that the pushblock happens. If you hit at least six buttons, the pushblock is guaranteed on that sixth button (if you hadn’t already gotten it). You have to press the buttons sequentially (you can’t just press all six attack buttons on the same frame and expect anything to happen).

In my experience, everybody tends to find a technique that works for them, and no two players’ techniques are identical. The method that works best for you will also heavily depend on your input device, but I promise that there are players who are consistently pushblocking on stick, on leverless, and on pad.

Usually pushblock methods come as some form of a slide input, a piano input, or some combination of the two. There are a ton of resources out there to help you learn how to do this. Generally, you want the first button you press in a pushblock attempt to be a button like LP or LK—that way if it turns out that you aren’t actually blocking anything, you get a pretty quick/noncommittal button.

When you’re just starting out, don’t worry about pushblocking in real matches! The important thing is finding a method you’re comfortable with and practicing it for a few minutes in training mode here and there. It’s easiest to practice against moves that put you in block stun for a long time, like Morrigan’s ES soul fist (236PP). Start working on the muscle memory early and it’ll become second nature in no time. If you’re having trouble finding a method that works for you, ask in the Vsav Discord server—there is an army of people who will help you!

3. Anti-Airing Is Probably Different Than What You’re Used To

In many 2D fighting games, grounded anti-airs are incredibly important. A jump that goes unchallenged leaves the attacker at a massive advantage. The developers generally give all characters grounded anti-airs that can consistently beat jumps. Any given Street Fighter character probably has at least one normal that will cleanly beat any jump-in, as long as you space and time that anti-air move correctly (and many characters will have specials that are even more reliable). You see the same thing in the Guilty Gear series, with characters’ 6P anti-airs having disjointed hitboxes to ensure they can beat aerial attacks.

Similar to the frame advantage stuff earlier, Vsav is the polar opposite of these games. In Vsav, in many situations, many characters will not have any grounded anti-air option that will work. Unsurprisingly, this makes Vsav a much more airborne fighting game than something like the Street Fighter series. That being said, there are plenty of answers to people jumping around; they’re just a lot more varied/nuanced than “stand in the right spot and press my AA button.”

When I’m thinking about how to deal with a jumping opponent in Vsav, the options I’m considering are usually some combination of:

1. Air-to-air: Often (but not always), the button your opponent will choose as a jump-in will be good against grounded opponents but weaker against airborne opponents. For example, wolf’s j.HP is a great air-to-ground button, but you can look to contest in the air.

   

2. Air-to-air, round 2: In Vsav, after blocking a move in the air, you can immediately press your own jumping button while still in the air. So instead of aiming to flat out beat their jumping button with your own jumping button, you can deliberately jump up and block their attack air-to-air, and then press your own button right after that to punish them. There is counterplay—your opponent can still chain normals in the air, so if they’re ready for this, they can try to jail you to the ground with an air chain.

3. Air-to-air, round 3: Continuing that line of thought, because you can chain normals in the air exactly like you can on the ground (LP > LK > MP > MK > HP > HK), you can aim to meet your opponent in the air and jail them to the ground with an air chain before they get a chance to press whatever jump button they were planning to press.

4. Don’t be there: Sometimes it’s best to avoid the button altogether if you can. This can come in the form of a walk-under/dash-under. For whatever reason, I find it a lot easier to walk or dash forward under jumping normals in Vsav than in other fighting games (and this is helped by cross-ups being much more rare here than they are in something like Street Fighter). This can be a great way to escape the corner!

5. Preemptive button: While challenging a jump-in as the attacker descends toward you may be difficult or impossible, you may be able to catch them on the way up. Attacks like Bulleta’s dash LP or Q-Bee’s 5MP are great at this. If you think someone is going to jump, but a button out there for them to jump into.

6. Pushblock: The only reason I’m putting this at the end of the list rather than at the beginning is because you probably won’t be consistently pushblocking when you’re first starting out. That being said, pushblock is the single most consistent anti-air option in Vsav. If you just stand there and pushblock a jump-in, you push your opponent back and gain some screen real estate.

The air game is incredibly important in Vsav, and it’ll probably take you a while to wrap your head around it. If you really want to play a grounded anti-air game, Felicia is the one character who can do that pretty reliably, so if you’re getting hung up here, you can always try picking cat.

4. Remember to Tech Roll

This is an easy one. After you get knocked down in Vsav, if you don’t press anything, you’ll get up in place. You can also press either left or right plus any button to roll either left or right. The timing isn’t strict at all (you can just mash buttons while holding left or right and consistently get the roll). Tech roll doesn’t have any punishable recovery and cannot be thrown.

When you’re first starting out, just randomly cycle between the three wakeup options (left/neutral/right). If you always choose the same option, your opponent can almost certainly run some oppressive oki on you, so by forcing them to try and guess or react to your roll, you’ll likely find yourself back to neutral more often than not after a knockdown.

5. Ireppa Is Your Friend

You’ll hear chicken block, upback, and ireppa all used interchangeably to mean holding some version of up to escape offense/pressure with a jump. While you can’t air block grounded normals in Vsav, holding up is still extremely strong. Jumping away may allow you to escape a situation or even start your own offense. If you play a character with a fast jump like wolf or cat, if you neutral jump, you may land in time to punish a whiffed button, or at very least, start your own offense by forcing your opponent to block your descending jumping normal. If you play a character with an air dash, you can aim to dash immediately after your jump to act even more quickly and put your opponent on the back foot.

If you’re worried about a high/low mixup, ireppa can be a good option. If they go low, they need relatively precise timing to catch your rising hurtbox as you jump. Most of the high attacks you’ll be worried about in Vsav are jumping attacks, so you can air block them. This means by holding up back, you’ll be doing a quasi-OS to avoid the low option while blocking the high option. Depending on your opponent’s timing, their high option might miss too.

A large portion of the cast has some kind of high/low mix that’s structured like this (threatening either a jumping normal or a low grounded normal). Force them to prove they can stop you from jumping away before you resign yourself to holding that mix.

6. Pick Whoever You Want

You might’ve noticed that pretty much everything we’ve talked about up until this point has been character-agnostic. Vsav exists in a sort of strange Goldilocks zone—it somehow simultaneously has characters that are incredibly distinct from each other while also having your Vsav IQ being relatively transferable from one character to another. When I first started playing Vsav, I played Bulleta for a bit before realizing she wasn’t for me and switching to Q-Bee. My time with Bulleta was not wasted—the understanding I’d gained of the game as a whole meant I was able to pick up Q-Bee that much faster. Longtime players can basically play random select at a high level because having solid understanding of Vsav means you can get up to speed with any character in the game in (relatively) short order.

7. There Are Great Resources Out There—Use Them

The two best starting resources for learning Vsav are the wiki and the Discord server. If you’re starting from absolutely square one, I also really like AmeriKame’s Vampire Savior for Dummies: A Beginner’s Guide.

Conclusion

I hope some of this was at least somewhat helpful. I honestly think Vsav is pretty easy to pick up in the grand scheme of things, and it’s a truly unique and amazing game. If there’s any interest, I could do a follow-up to this post, but in the meantime, get out there and play some Vsav!

The beginner guide for MARVEL Tōkon: Fighting Souls just dropped. I’m just going to refer to it as Tōkon unless someone comes up with something better. I thought it'd be fun to break down the guide to better understand how the game might play. I also like having a record of how the game looks at this moment in time (there will inevitably be changes between now and release). Finally, it’s just fun to try and predict how the game will turn out (I assume I will be wildly wrong, but it’s my blog, and nobody can stop me).

The Control Scheme

When I’m trying to wrap my head around any new fighting game, the first thing I want to know is what buttons exist—I think there’s no better way to start understanding the game. Tōkon has a bunch of buttons:

• Light Attack (L)

• Medium Attack (M)

• Heavy Attack (H)

• Assemble (A): Your assist button

• Unique Attack (U): Similar to the drive button in Blazblue or the stand button in JoJo’s Bizarre Adventure (your “do your character’s unique mechanic” button)

• Quick Skill (QS): A button for one-button special moves

• Quick Assemble (QA): Does a variety of different team actions—we’ll talk about this one later since it doesn’t have a good point of comparison from a past fighting game that we can use

• Quick Dash (QD): Dash button

One reason I start with the control scheme is because one of my earliest goals is figuring out what my button layout will be. Part of that question is how many of these buttons are actually necessary. L, M, H, Assemble, and Unique Attack are the core five buttons for this game, and they’ll be needed no matter what, so at minimum, this is a five button game.

The three “Quick” buttons could all theoretically be omitted. Quick Skill lets you do your special moves without doing motion inputs. Motion input specials do a bit more damage, so you’ll want to do those when possible (in combos/when reaction times don’t matter), but as we’ve seen in Granblue Fantasy Versus, there will be times where in high level play, you’ll want to use the one-button specials—in situations where you’re trying to react to something, it’s much easier with a single button press, so I think Quick Skill is functionally mandatory.

You could also technically do without Quick Dash, since you can also dash by double tapping a direction, but pressing a single button to dash is unambiguously better than double tapping a direction, so this button is mandatory as well.

Quick Assemble can also be done as a single input by doing L plus the Assemble button, so this is the one button that can be safely omitted.

Putting all of this together, we’re looking at a seven button game. When thinking about my button layout, I’m also thinking about what buttons might need to be pressed together. Like we just talked about, we’ll be doing Quick Assemble by pressing L+A, so we’ll want those two buttons near each other if possible. You do throws with M+H, so we’ll want those buttons near each other as well. I’ll want the dash button in an easy to reach place, and to the extent that it’s possible, I’ll want the layout to be similar to what I use for games I’ve played in the past, to minimize the learning curve.

For me personally, I’m leaning toward:

We’ll talk more about this later!

User Interface

After looking at the controls, the next thing I look at when trying to understand a new game is the UI during a match. If the developer thinks something is important enough to represent it graphically on screen at all times, it’s probably important to know what it is.

Shared Life Bar

This was hinted at before this video dropped, and now we have confirmation that rather than individual characters having their own life bars (and being able to be taken out individually), your team has a collective life bar. This will help avoid slippery slope situations like I talked about in a recent video, but it also means if the opposing team has an assist that’s hard to deal with, you can’t ever kill that assist character—you just have to deal with that assist forever. That being said, hitting the assist does do damage to the collective life bar, so you can still punish opponents for poor assist calls.

Character Portraits

Next to your health gauge, you have the four portraits of the characters each player picked for their team. When the match starts, you only have access to two of your characters (presumably the first two you picked on the character select screen) and the other two are greyed out. When you meet certain conditions, you unlock your third and fourth characters, giving you access to their help via the game’s different team/tag mechanics that we’ll talk about later. Unlocking additional characters also increases your total health and maximum possible super meter:

Round Counter

Versus games traditionally don’t have rounds like you see in most 1v1 fighting games. Tōkon differs from them and has a best of five round format. After losing a round, you unlock your next character for the remainder of the match (if you have yet to unlock all of your characters).

Traditionally in versus games, the losing player will have access to fewer and fewer mechanics as the match goes on (since they are losing characters). Tōkon flips that on its head and instead has the winning player gain access to more and more mechanics as the match continues. In some ways this is functionally the same. Whether it’s the winning player becoming more powerful or the losing player becoming less powerful, you’re still creating a slippery slope situation. It differs in that:

1. The losing player will still eventually unlock all their characters, so the advantage given to the winning player is temporary and

2. You can never have less than two characters, so you will always have access to team mechanics (unlike a traditional versus game where you lose access to them completely when you’re down to your final character).

There’s also the psychological phenomenon called loss aversion—the pain of me giving you $20 and then taking it away feels worse than me just not giving you $20 in the first place. Giving the winning player more stuff will feel better than taking away the losing player’s stuff.

Assemble Gauge

It’s a bar with three chunks underneath your health gauge that’s used for certain team mechanics.

Skill Gauge

You spend this to do supers and EX moves.

Basic Mechanics

Basic Attacks

Light, medium and heavy attacks seem like they work exactly like they do in Dragonball FighterZ, down to each button having its own auto combo and 2H being a universal launcher.

Assists (“Assemble”)

Like other versus games, you can call in your off-screen characters to help while still controlling your point character. Tōkon handles this a bit differently than we’ve seen in the past. There’s just one assist button (they call it “Assemble,” but I’m going to talk like a human here and just call them assists). You choose the assist you call by pressing a direction plus A. You call assist 1 by doing 5A or 6A, you call assist 2 with 2A, and you call your third assist with 4A.

Input-wise, I don’t love this, for the same reason that I don’t like needing directionals for assists in 2XKO. It makes it impossible to do things like block while calling assist 1. This is why I’m considering adding directional buttons to my control layout so that I can call the assist I want more easily:

Tōkon is also doing a few odd/new things with assists.

First, the assist a character does is dictated by their character slot. The character in assist slot 1 will do a projectile assist. If that same character gets swapped to assist slot 2, they’ll do an anti-air whenever you call them as an assist. If they end up in slot 3, they do a unique character-specific assist. This might not be as hard to keep track of as it first seems. Assist 1 will always be a projectile (or something projectile-esque), and assist 2 will always be a vertical/anti-air assist. However, it does seem like this might just homogenize the cast and make team comp boring, since each character’s assists might be largely interchangeable (outside of the slot 3 character-specific assist). If projectile/anti-air assists do differ a lot from character to character though, that homogeneity could be mitigated (at the expense of making the game a lot more confusing).

Second, you get a different assist entirely if your opponent is in hit stun or block stun, so it seems like every assist just becomes a combo/pressure extender when used while you’re hitting your opponent. This use of an assist also costs a chunk of assemble gauge, presumably to prevent infinite combos and indefinite pressure sequences.

Calling an assist in the air does an aerial version of the assist which seems to track to your height when you called the assist. This could make certain types of aerial strategies impossible—you may not be able to call an assist on the ground while flying around in the air.

You can also swap to playing your assist character by tapping the assist button during their attack or immediately after their attack. This seems a bit more restrained than you see in the active tag systems in games like Marvel vs. Capcom: Infinite or 2XKO. I’m a fan of this approach since active tag in those other games can sometimes turn into feeling like you’re forced to play two characters simultaneously for a lot of the match. It can also make things visually really hard to keep track of.

Special Moves, EX Special Moves and Supers (“Skills”, “EX Skills”, and “Super Skills”)

Again, I’m going to stick with human language and just call these specials, EX specials, and supers. They aren’t really doing anything new or crazy here.

Specials can be done with motion inputs or with the Quick Skill button. As we talked about earlier, there will be times where either the motion version or the one-button version will be best, so you’ll want to have both in your arsenal (see: Granblue Fantasy Versus).

EX specials cost one chunk of super meter (25 meter) and allow you to do a powered up version of that special (similar to how they handled EX moves in Dragonball FighterZ).

Supers cost 50 meter, and level 3 supers (“Ultimate Skills”) cost 100 meter.

Team Mechanics

Assemble Smash

Costs 1 Bar of Assembly Gauge

On the ground, Assemble Smash is a mix of a universal overhead and drive impact from Street Fighter 6. You press the Quick Assemble button (or Light Attack+Assemble) to do an armored (seems like hyper armor?) advancing overhead attack. The startup is around 27 frames, so it should be pretty easily blockable on reaction. I assume this will be a popular YOLO option in the early days of the game. It’s not clear at this point how safe it is on block.

The air version is a bootleg DBFZ super dash, where you advance forward with an attack while nullifying projectiles. It’s hard to tell how strong this is. We’ll need details like:

1. How far it can go (does it go a fixed distance?)

2. How much recovery it has

3. Whether there’s any kind of vertical tracking

Assemble Counter

Costs 1 Bar of Assembly Gauge

Assemble Counter seems most similar to V-Shift from Street Fighter V. You press 4+Quick Assemble (or 4+Light Attack+Assemble) to do a back dash that will counter attacks if they make contact with you as you move backwards. It’s punishable during recovery, so you can probably be blown up pretty badly for overusing this.

Super Assemble

Costs 1 Bar of Assembly Gauge plus 50 Skill Gauge

This is just a DHC. It’s done by pressing Assemble during a character’s super. You can do up to two in a row.

Crossover

Costs 1 Bar of Assembly Gauge

Done by pressing Light Attack+Assemble while blocking, this is an alpha counter with a hint of Xrd blitz shield. If your opponent doesn’t do anything, you get a knockdown and get pressure afterward. However, your opponent can react with their own Crossover (they call this Crossover Reflect). If they do this, it seems to put you back to neutral. If you do Crossover Reflect with the right timing though, instead of resetting to neutral, you knock them down and get offensive momentum.

The possible sequences here are:

1. Player 1 is blocking > Player 1 does a Crossover > Player 2 doesn’t react > Player 1 is rewarded with offense.

2. Player 1 is blocking > Player 1 does a Crossover > Player 2 reacts with Crossover Reflect > Players are reset to neutral.

3. Player 1 is blocking > Player 1 does a Crossover > Player 2 reacts with a perfectly timed Crossover Reflect > Player 2 is rewarded with offense.

How this actually plays out will depend on how difficult it is to react to Crossovers and how difficult the just frame timing is when you do a Crossover Reflect. If it’s difficult to react to Crossovers at all, offense will feel very weak. If players get to the point where they can reliably do perfectly timed Crossover Reflects, Crossovers will become garbage.

There’s still a lot we don’t know here. Can Crossovers be baited with quick attacks that recover in time to block the Crossover? How real is the offense you earn from a successful Crossover or a perfectly timed Crossover Reflect? We’ll have to wait and see.

Conclusion

I don’t know that I have a coherent way to end this. If I missed anything or made any errors, please let me know. Before the starter guide video dropped, I was cautiously optimistic about the game. After watching the video, that’s still pretty much where I’m at.

When I started work on my video about cursed design problems in versus games, it was originally going to be very different. Instead of the video it became, it was going to be about how versus games have strayed away from the formula we saw really crystalize in the Marvel vs. Capcom games. I still wanted to talk about it, so here it is!

My History With Marvel vs. Capcom 3

I’m someone who played fighting games as a kid in the 90s… extremely badly. I enjoyed playing arcade mode in Street Fighter II Turbo and Killer Instinct on my SNES in my bedroom (hooked up to my own personal 13” Toshiba CRT TV). I had no idea there was a community of people who played these games competitively.

Marvel vs. Capcom 3 was the first game where it all came together for me. It came out at the right time and place—I had free time, I had a local scene, and I loved the game. From when the game came out in 2011 until I moved at the end of 2016, Marvel 3 was the thing I did in my free time—I loved it. I entered locals, helped my buddy start a weekly that he ran out of his apartment, and managed to get out of pools at EVO one year.

Even before I moved away from my scene though, it was pretty evident the game had lost some momentum—sooner or later there will be players and viewers who want to move on to something new. The game’s awful netcode was also a major factor in its decline. A game with good rollback netcode can pretty easily maintain an active playerbase online. If you can play good matches with anybody on your continent, you can all be pretty spread out, and there doesn’t need to be that many of you. But without good netcode, if the scene shrinks to the point where you can’t consistently get offline play where you live, the game is functionally dead (for you personally at least). People in a few areas had the player density to keep the game going, but in lots of places, it disappeared. People did figure out how to play online using Parsec instead of using the netcode built into the game. It’s an improvement, but it’s still a far cry from proper rollback netcode.

If I could’ve kept playing Marvel 3 forever, I probably would have, but I moved, and the game was gone (for me). I did still love fighting games, so the search was on for something else to play. I picked up Guilty Gear Xrd, but it never quite filled the Marvel-shaped hole in my heart.

Fast forward a bit to mid-2017. We had three games on the horizon all vying to be a successor to Marvel 3:

• Marvel vs. Capcom Infinite (released September 19, 2017)

• Dragon Ball FighterZ (released January 26, 2018)

• BlazBlue: Cross Tag Battle (released June 5, 2018)

MvC:I and Cross Tag Battle both faded away pretty quickly (for reasons beyond the scope of this post), and DBFZ quickly stood alone as the post popular versus game from that era.

At a glance, DBFZ is very similar to Marvel 3. You pick a team of three characters, and then you pick one of three available assists for each character. The three characters share a common resource (super meter). Both games have various ways to tag in other members of your team with different risks and costs. Some damage that characters take can be recovered if you tag them out and let them heal.

DBFZ clearly patterned itself closely after Marvel 3, but it never really felt the same to me. This isn’t a slight to the game—I’ve played DBFZ relatively consistently since it came out (I wouldn’t be playing it if I hated it!), but as somebody who wanted more Marvel 3, DBFZ was (in many ways) not that. It was something that I think I struggled to wrap my head around and articulate for a while, but it’s been at the back of my mind every time I’ve booted up DBFZ.

A few things got me thinking about this again recently. Marvel 3 was brought back for EVO 2023 as a throwback title (having not been at EVO since 2017). It beat DBFZ’s entrant numbers pretty handily:

And of course we are lucky to have dedicated communities who have been keeping the game alive this whole time (shout outs to Tong!)

Around this same time, a player made a post on the Marvel 3 Subreddit titled, “I’m a DBFZ player considering getting into MvC3.” The years of random thoughts about how the two games differed crystalized in my mind, and I typed out a big list of how the games were in many ways almost exact opposites. I thought it’d be fun to expand on those thoughts a bit and put them somewhere easier to find than a random Reddit comment.

Difference 1: Assists

Assists are the defining characteristic of games like Marvel 3 and DBFZ. They play the single biggest role in why the unit of play is a team, rather than three separate characters you cycle through playing. It’s why Marvel 3 and DBFZ are fundamentally different from The King of Fighters. Because the unit of play is your entire team, and changing a single assist or swapping out a character can wildly change the dynamic of your team, assist-based fighting games let you “customize” your character (i.e., your team), since there are thousands of possible character/assist/team order combinations which are all meaningfully different (I made a whole video about this!)

If assists are the defining characteristic of both Marvel 3 and DBFZ, the differences in how assists work in each game will help start to explain how the games differ as a whole.

In some ways, assists in DBFZ are stronger than assists in Marvel 3. In Marvel 3, you can’t call assists when your point character is doing a special move or if you super jump, while DBFZ lets you do both of these things. Calling an assist in Marvel 3 locks out your other assist for a short duration, but DBFZ doesn’t have this restriction (you can call both assists at the same time if you want). Also, compared with Marvel 3, DBFZ assists also tend to cause more block stun and a ton more hit stun. Turning a stray assist hit into a combo in Marvel 3 can be challenging, but in DBFZ, it’s pretty easy.

Despite all of that, I’d argue that if you look at the strengths and weaknesses of assists in the two games and compare them, assists are both stronger and much more important in Marvel 3. The primary reason for this is assist cooldown, which dictates how often you can call your assists.

In Marvel 3, after you call an assist, both assists are locked out for a bit. The lockout for the assist you didn’t call always ends 30 frames before the lockout for the assist you did call (so you can call your assists slightly more often if you alternate between your two assists). The lockout time varies a bit from assist to assist, but it ranges from around 2.5 seconds for quicker assists, and 3.5 seconds for slower ones. It varies from assist to assist in DBFZ also, but regular assists have a cooldown closer to 8 seconds (for the quickest recovering assists).

DBFZ also has a bunch of restrictions on when the cooldown timer will even begin to start counting down. If your assist is involved in a combo, your assist cooldown timer doesn’t start ticking until after the combo is over. If your assist is used in a blockstring, it won’t begin recovering until after you have deliberately left a significant gap in your blockstring.

In Marvel 3, you can call assists way more often than in DBFZ. However, it comes with the meaningful downside: it is much easier to attack and kill assists in Marvel 3. Assists are so strong that you need to use them to play the game at a remotely competitive level, but because they are so vulnerable, a poor assist call can cost you a character (or two). You can base your entire strategy around fighting your opponent’s assist instead of their point character, which is cool as hell and mostly absent from DBFZ.

Difference 2: Zoning

Zoning in DBFZ sort of exists. Lots of characters have very quick low commitment projectiles with their ki blasts. You can spend a bar of super meter to turn any stray hit into a combo, so any random ki blast can become an easy confirm into a combo, and you can turn that combo into continued offense/pressure. On the flip side, characters have universal tools to deal with projectiles. Super dash goes through ki blasts completely, and while the character throwing the ki blast can anti-air the superdash if they’re ready for it, the existence of super dash is a strong deterrent to throwing ki blasts at certain ranges.

Some projectiles (like beams) cannot be superdashed through, but these tend to be very committal, so the character who avoided the beam can often punish the opposing character. Additionally, while it costs a bar, vanish allows you to instantly teleport behind the opposing character and hit them, so as long as you have meter, you always have a way to bypass zoning.

Finally, chip damage in DBFZ does not matter. You can block projectiles until the clock runs out. I tested it with Goku Black’s beam, and to actually do a character lifebar’s worth a chip damage, it was around 155 blocked beams. Also, all chip damage is recoverable, so you can tag the character out and heal completely. You can’t die from chip damage, so even after 155 blocked beams, they still have to hit you to actually kill you.

Zoning in Marvel 3 is good. Really fucking good. There aren’t any universal mechanics like super dash or vanish which can bypass zoning, so you have to deal with it the old fashioned way. You have to either punch through with your own projectiles or slowly work your way in. If you haven’t seen it, watch the Curleh Mustache 2 grand finals. There is no equivalent to this in DBFZ.

Characters like Morrigan can almost completely fill the screen with projectiles. Characters like Magneto have extremely quick low risk projectiles with his beam and magnetic blast. Because you can call assists much more frequently in Marvel 3, on top of whatever your point character is doing, your assist can also constantly be controlling the screen.

Chip damage in Marvel 3 is substantial. Instead of 155 beams in DBFZ, you can take out the average character's health bar with ~18 blocked beams. In other words, chip damage in Marvel 3 is an order of magnitude higher than it is in DBFZ. Unlike DBFZ, some chip damage is permanent damage which cannot be recovered, and you can also die to chip damage. Magneto with Doom beam assist can chip out an average health character in ~20 seconds if they just stand there. Doing the same with Goku chucking beams backed with two beams assists in DBFZ took me three minutes and twenty seconds.

Difference 3: Character Mobility

Character mobility options in DBFZ are near-universal. Every character can run, air dash backward and forward, super jump, and double jump. Every character’s run speed, air dash, super jump, and double jump are the same. There are a few exceptions, like SS4 Gogeta with a double air dash or Zamasu with flight, but for the most part, the entire cast relies on the same set of universal mobility tools.

Character mobility varies wildly in Marvel 3, where there is arguably more variety in its air dash options across the cast than the entirety of mobility variety in DBFZ. Some characters can wall jump. Some characters can double jump. A few characters can triple jump. Some characters can fly. Some characters have teleports that all function in different ways. Arthur can’t dash at all. MODOK can’t normal jump. Some characters have unique movement options (like Spencer grappling hook or Spider-Man web glide). Some characters can float. Almost no two characters navigate the screen in the exact same way. Characters feel much more distinct when compared with the relative homogeneity of DBFZ movement options.

Difference 4: Verticality

This kind of goes together with the character mobility differences, but Marvel 3 is a much more vertical game than DBFZ.

For most of a DBFZ match, you can generally see the ground. There are rare exceptions like if someone drops a superdash combo at the top of the screen after a 2H or something, but most of the game takes place in a vertical space that’s pretty similar to most 1v1 2D fighting games.

In Marvel 3, the screen is as tall as it is wide—you can spend a lot of the match as far vertically from your opponent as you are horizontally. 10 seconds could pass without you seeing the ground.

The verticality makes assists more interesting as well. DBFZ and Marvel 3 both have vertical beam assists (UI Goku rising heat and Dante jam session), but in DBFZ, the verticality of the UI Goku’s assist just matters a lot less because you’re probably not that far off the ground in the first place. In comparison, jam session creating a vertical wall that reaches the top of the screen matters a lot in Marvel 3 because your opponent might be miles above you. More of the screen matters in Marvel 3, which means there is a greater variety of space to control, and that matters for both point and assist characters.

Difference 5: Defensive Options, Pressure, and Mix-ups

DBFZ’s unique defensive mechanic is reflect. If you time it correctly, you’ll stop your opponent’s attack and push them back. If they committed to another move, that move might now whiff completely, allowing you to land a hit of your own. Reflect is risky though. You’ll often have to reflect as a read (rather than a reaction), and a missed reflect is very punishable if your opponent is ready for it.

Marvel 3’s push block (officially called advancing guard) seems similar at a glance. Push block gets you some breathing room, just like reflect does in DBFZ. The details matter though. You activate push block when you are already blocking a move (rather than guessing when your opponent will do a move like with reflect). Reflect is a high risk/high reward defensive option to call out and possibly punish a high commitment move. Push block is a low risk/low reward mechanic that pushes the game back to neutral, preventing hours-long pressure sequences or blockstrings.

You can’t talk about defense in a bubble, and in terms of pressure and mix-ups, DBFZ and Marvel 3 feel diametrically opposed. In DBFZ, it’s relatively easy to put your opponent into blockstun. Plenty of moves jail into vanish, so if you can get your opponent to block your beam at full screen, you get a guaranteed vanish and now you’re +2 at point blank range. While you can get next to your opponent and start a blockstring with relative ease, mix-ups are generally weak without specific blockstun assists to force an actual 50/50. It can be hard to shake the feeling that DBFZ is just the players just taking turns blocking and dealing with reactable offense.

In Marvel 3, it’s relatively hard to get your opponent to hold a grounded blockstring. The strength of push block (and chicken blocking) make it hard to keep any kind of sustained pressure. It’s not that hard to slip away and reset to neutral. However, mix-ups can be truly absurd when you do manage to earn one. Throws are one frame and can lead to a dead character. If a player gets stuck blocking Amaterasu cold star assist on the ground, god herself is not blocking Magneto tap dancing on her head.

DBFZ is a game where it’s easy to find pressure, and that pressure can go on for a while, but mix-ups are weak. Marvel 3 is a game where it’s difficult to find pressure or force mix-ups, but the reward is enormous when you do.

Difference 6: Losing at Character Select

I touch on this the Cursed Design Problems video, but I’ll briefly talk about it here. I think the synergies that make your characters cohere together as a team are what makes versus games truly special. If you can pick any random three characters and it works alright, it probably means synergies just don’t matter all that much. DBFZ letting any random three characters work as a team does mean that people can just pick their favorite characters (and it’s understandable that the devs want to give people this freedom), but I feel like some magic is lost here.

Let me give a goofy example from Marvel 3. I made up a team of Hawkeye/Rocket Raccoon/Morrigan. The plan was to zone with Hawkeye while calling Morrigan dark harmonizer assist to build meter. Then as soon as I got two meters and was under any kind of threat, I would super with Hawkeye and DHC into Rocket’s mad hopper super. Hawkeye super is very hard to challenge, and then mad hopper sits there as a trap for ages that you can just park your character on top of. Finally, I’d raw tag Hawkeye back in, and have him hang out and zone on top of the trap while calling Morrigan assist to hopefully do the same thing again. This didn’t turn out to be a particularly amazing team, but the theory of the team was cool and it was fun to mess around with. I never found anything that quite felt like that in DBFZ. In Marvel 3, theorycrafting was almost as much as actually playing the game.

Difference 7: Confirms and Combos

This is a minor point, but I liked the difficulty of hit confirms in Marvel 3. Some were really easy of course, like Doom foot dive or Haggar pipe, but there were confirms that were kind of hard, but doable enough that they were absolutely worth going for while being absolutely droppable. Watching RayRay turning stray hits into proper combos is amazing. In DBFZ, you can just always fall back on, “I’m not sure what to do here, guess I’ll vanish,” and boom, your combo continues.

Marvel 3 also feels like you need to optimize combos reasonably well to play the game at a high level—combos need to lead to a kill. Combos that kill a healthy average HP character might be reasonably droppable, so you often have important decision points of whether you should transition to an easier combo if the full combo won’t be necessary (if the opposing character is close to death anyway).

Between lower execution requirements and the smaller difference between decent and truly optimal combos, this sort of thing just doesn’t come up in DBFZ nearly as much. It feels like I usually need two touches to kill a character, and it often doesn’t much matter if I do an optimized combo or a day 1 bnb.

Difference 8: Tagging and Resource Management

In Marvel 3, meter management is crucial. Raw tagging to swap characters is wildly risky, so the most common way to safely tag is via DHC, which costs two meters. Having meter on deck for a safe DHC is critical. In DBFZ, there are other reliable ways to tag characters, so safe DHCs don’t really matter much.

You can also spend meter in Marvel 3 for utility in interesting ways. Whether Spencer has a bar (and therefore has access to bionic arm) changes the match completely. The same goes for Dormammu sitting on bar for chaotic flame, or Doom/Amaterasu sitting on two bars for a THC. Meter dictates the match in a lot of cases. In DBFZ, this feels less important, and meter use for utility mostly exists in the form of EX moves (apparently these are called powered-up specials?) and vanish. Supers are mostly for damage after you’ve already landed a combo.

Closing Thoughts

As I got further down in the list, I probably sounded increasingly less evenhanded and more antagonistic toward DBFZ. Ultimately, the point I’m trying to make is that DBFZ are very different once you get past surface level stuff.

I think that I could do a version of this for most other post-Marvel 3 versus games (DBFZ just happens to be the one that I’m the most familiar with). I’m sure that Marvel 2 vets could just as easily break down why Marvel 3 was not a proper successor to Marvel 2. Everybody is chasing their own white whale.

Getting into retro fighting games gave the whole genre a second lease on life for me, so I wanted to try to convince other people to give them a shot. In one way or another, a lot of my videos are (not so secretly) just me trying to get people to like the stuff I like and play the stuff I play (don’t tell anybody!). If you haven’t seen it yet, I recommend watching the video before proceeding:

All of my videos come with caveats, but I’m fortunate that you all give me some grace and assume good intentions on my part (even if you disagree with certain things, which is absolutely fine!) That being said, this video comes with even more caveats than usual, so I wanted to talk through some of the pushback I think it’s likely to get.

For every single chapter in the video, I think there are some folks who will strongly believe the exact opposite thing I say. My imagined conversation goes something like this:

Me: Having no patches is great! Everything I learn will be in the game forever, and I don’t have to keep up with constant updates.

Reasonable Person Who Disagrees: Having no patches sucks. Good characters stay good, and bad characters stay bad. The game is just less balanced in general.

Me: A small cast is great! I can really dive in and learn about each character and have a solid understanding of every match-up.

Reasonable Person Who Disagrees: Having a lot of characters is fun. I like variety and I like having a lot of characters to choose from in a bunch of different archetypes.

Me: Old games that have survived for 20 years aren’t going anywhere, and that’s great! I can keep playing the game as long as I want.

Reasonable Person Who Disagrees: I don’t need something that will last for the rest of my life. I want to play for a few months or a few years or whatever, and I can be 100% happy with that. For now at least, thousands or tens of thousands of people are playing this game, and it’s fun playing something new and being part of something so big.

Me: Old-timey fighting game communities are great! People are super nice and welcoming!

Reasonable Person Who Disagrees: Sure, but new games have nice communities too. I go to my local and play Street Fighter 6 with people I genuinely enjoy hanging out with.

I 100% understand every one of those counterpoints—they’re reasonable! In the video, I’m not trying to say that they aren’t valid points of view. Instead, I’m trying to point out that I think those counterpoints are held as the conventional wisdom—universal truths that people accept, and this scares people off from even thinking about trying old-timey games. I’m just trying to say it’s mostly a perspective thing. Anyway, if you want to play some Vampire Savior, hit me up!

Other Random Thoughts That Didn’t Fit Anywhere Else

• The “No Patches” chapter kind of touches on why you don’t see too many efforts to rebalance old games, and why rebalanced versions rarely stick around or find much of an audience. People have invested literal decades into learning some of these games, and they’re understandably reluctant to throw that away.

• I was originally going to have a whole chapter just about why Fightcade is dope, but I decided against it (though I do mention Fightcade a bit in the “Community” chapter). I thought it’d simply make the video too long and it was mostly unnecessary—anybody looking to play VSav or 3S is going to end up on Fightcade whether I direct them there or not. That being said, I really do think Fightcade is amazing and makes the retro fighting game experience something even more special.

Most of my posts are about fighting games themselves, but this one is about my attempt at building my ideal arcade stick. My goal was to find the best button layout possible (for me, at least). If you’re interested in my thought process or prototyping your own layouts, you’re in the right place!

Disclaimer: I’m only addressing prototyping layouts in this post. I’m not going to talk about PCBs, wiring, etc. (though I can do that in a future post if anybody is interested).

A Brief History of Retail Arcade Sticks

Commercial arcade sticks usually use one of a handful of popular arcade cabinet button layouts. I bought a Mad Catz TE1 back in 2009 which uses the layout from Vewlix arcade cabinets. Mad Catz had the license to make official Street Fighter IV peripherals, and because SFIV was on Vewlix cabs, Mad Catz used that same button layout on their sticks. Mad Catz dominated the arcade stick market from the SFIV era up through the early PlayStation 4 days (2009 up through 2016 or 2017), and their success at that time helped make the Vewlix layout the most common one you’ll see on retail sticks. I played on the Vewlix layout for well over a decade and never really questioned it.

The Cross|Up

Hit Box Arcade announced a Kickstarter for a new product called the Cross|Up in 2019. After it was officially released, curiosity eventually got the better of me, and I bought one in 2022 (I wrote a whole post about it!) Just like in that post, I’m just going to write the name as “cross-up” from here on out. I really ended up liking it, which eventually got me thinking—if I like this layout better than Vewlix, is there an even better layout out there for me?

In my post about the cross-up, I mostly focused on the utility of the extra buttons around the main button cluster.

It's also interesting to look at the main cluster itself and see what they did with it. In the picture below, I've overlaid the main action buttons from the cross-up layout (in pink) on top of the Vewlix layout (in green). You can see the cross-up layout is quite a bit more compact:

The tighter button spacing made it a lot easier to reach the additional buttons. After getting used to the more compact cross-up layout, I couldn’t go back to Vewlix anymore; it felt like the buttons were miles apart.

Improvements

I think it's pretty reasonable to expect that what's best for me (layout-wise) won't be the best for everybody, but I thought it’d be fun to go through my personal hunt for the perfect layout.

I focus on 2D fighting games like Ultimate Marvel vs. Capcom 3 and Vampire Savior. Marvel has four main action buttons: L, M, H, and S. Having my index finger, middle finger, ring finger, and thumb on those buttons lets me access all of them without having to shift my finger/hand position at all:

It seemed like it would make sense that this would be a good starting point when searching for my layout. I would put four buttons where those four fingers naturally sat when my hand was in a relaxed/comfortable position, and then I could build on that.

When manufacturers started adding a seventh and eighth action button to arcade stick layouts, they almost universally added those buttons to the right:

Fighting games generally required six buttons at most, so for home ports of fighting games, you might end up using these extra buttons for training mode functions like recording the dummy or resetting your screen position. If you do want to use these for actual action buttons though, they can be a bit hard to reach. Your pinky can kind of hit the upper right button, but even then, it's your pinky, the most garbage of the fingers. I always wondered why they didn't put buttons on the left:

You'd have to shift your hand a bit to hit them, but you'd be using two much stronger digits (your thumb/index finger). Also, because you've got those two fingers right there, you could easily hit the two buttons simultaneously if you wanted.

In addition to adding buttons in reachable spots, the more compact you make the button spacing, the easier it is to reach any given button. With the cross-up, they stuck with 30mm buttons, apparently because they were (probably rightly) concerned with not wanting to deviate too far from what players were used to (basically all arcade cabinets and commercial sticks use 30mm action buttons).

I wanted to experiment with 24mm buttons. I thought there was a pretty decent chance I'd find a 24mm layout to be too small/cramped, but I figured it was worth trying (and folks have been using 24mm buttons on leverless layouts for ages).

To recap, my goals were:

1. Tailor button locations to my hand size and use case.

2. Add more buttons in easy to reach spots.

3. Make the layout as compact as possible while still keeping it comfortable.

Prototyping

My first attempt was a rickety enclosure made out of MDF and some scrap wood. I figured out a layout to try using Adobe Illustrator, printed it out, and then I taped it to a piece of MDF and went to town with a hole saw. Unfortunately, while I snapped a few photos during various points in this process, I wasn’t doing it with a future blog post in mind, so I don’t have the best visual aid here (sorry about that).

When doing the super early prototyping (the first two or three attempts), I knew within five minutes of trying the layout that there was a bunch of stuff wrong with it, and I immediately needed to make a new panel to fix the issues. This means at this stage, whatever you do, I highly recommend you do it as cheaply and quickly as humanly possible. Find some way to build a box, and cut holes in a panel on top of that box. It doesn’t need to be built to last at this point.

After I got past the major obvious layout issues, I was getting to a point where cutting panels by hand with a hole saw just wasn’t accurate enough—it got to the point where shifting buttons around by a few millimeters in one direction or another was important, so I abandoned the approach.

Laser Cutting and The Standoff Box

My solution to the accurate cuts problem while still keeping things cheap was ordering laser cut hardboard from SendCutSend. You can upload a vector file (I used Illustrator and just used the *.ai file format) and they’ll use it to cut your panels. I just made a top and bottom and then separated the two pieces with metal spacers:

Enclosure cost breakdown:

• Top and bottom panels from SendCutSend: ~$38

• Standoffs from Aliexpress (I used the 60mm M6 ones here): ~$15 (we need eight of them)

• M6 screws from Aliexpress (I got 10mm long ones from here): ~$3 (we need 16 of them)

• I play on a lever, so I also needed:

  • M4 screws from Aliexpress (I got the 16mm long ones here): ~$2 (we need four)

  • M4 bolts from Aliexpress (I got these) : ~$1

  • Spacers from SendCutSend to mount the lever: ~$4

So the total cost for all of that ends up being around $65. I don’t think that’s doing too bad! Obviously you can reuse the hardware as you prototype, so each subsequent iteration costs you $30 (since SendCutSend has a $30 minimum). $30 per new layout to test isn’t nothing, but the resulting enclosure was surprisingly sturdy. I had no issues playing on it for weeks and it was rock solid the entire time.

I’ve uploaded the Illustrator files to Google Drive (links at the very bottom of the post) if you want to edit them and try out prototyping like this.

Making the Nice Version

Once I was far enough long that I thought I was an iteration or two from a finalized layout, I bought a 14” modular case from AllFightSticks. I didn’t order a top panel, so this ended up being about $160. I was planning on using this for the final version of my stick that’d I’d play on for years, so I figured I would get the case, and then zero in on my final layout with custom panels from SendCutSend. I tried three more layouts in different materials before finalizing it.

After I was certain I was happy with the layout, I ordered a matte acrylic panel from AllFightSticks just because I liked the look/feel of it (though I would’ve been fine playing on the SendCutSend panel forever). Here’s the final product:

Random Advice

Here’s a random list of stuff that I wasn’t sure what to do with:

• Make sure your button spacing isn’t too close! Different buttons have slightly different diameters. I found that making sure buttons were at least 5mm apart meant that most buttons would fit fine. If you want the most compact layout possible, Sanwa OBSF and Qanba Gravity KS buttons have smaller diameters to accommodate this. Others, like Punk Workshop or Hori Hayabusa buttons are larger, so if you want to use those, measure carefully.

• Make sure your button hole sizes and screw holes are correct.

• Before placing a SendCutSend order, print out the layout at home. See what it looks like on paper and avoid wasting money if there are big/obvious mistakes.

• If you’re doing the standoff box method, don’t forget spacers so that the lever is recessed below the panel and sits at the right height.

Final Thoughts

You’re going to have to prototype a lot, so find a way to do it cheaply (but still accurately), especially after you progress past your initial terrible designs)

If you’re deep into fighting games and have the interest, time, and money to do this, I think designing your own layout is absolutely worth it. If you’re going to spend three or four or five hundred dollars on a custom stick, it seems weird to default to a Vewlix or Sega 2P layout without even questioning it and actually knowing it’s the layout for you.

Files and Resources

Standoff Box

These are the files I sent to SendCutSend to prototype button layouts. I sandwiched 60mm standoffs between the top and bottom panels which allowed enough room for wiring/buttons (50mm would likely work as well, but I did not test this). The lever spacing plate is needed to recess the lever beneath the top panel so that it sits at the correct height (I used two of them). The panels I ordered were all 0.125" hardboard.

• Standoff Box files for SendCutSend

Hardware:

• 60mm M6 standoffs x8

• 10mm M6 screws x16 (to attach the top and button panels to the standoffs)

• 16mm M4 screws x4 (to attach the lever)

• M4 bolts x4 (again, if you're using a lever)

Final Panel Design

• Final top panel file for AllFightSticks 14” case for SendCutSend

I finally found the motivation to put out a video (go check it out!), and I still can’t decide if it makes sense, if anybody will agree with it, or if my analogy between the handbook idea and my thesis holds up to any amount of scrutiny.

Random Caveats

I’m not advocating trying to simultaneously learn multiple games. Any fighting game requires a good amount of dedication/focus to make any headway. I’m just saying over the course of years and years, I think it’s helpful to play more than one game. As an example, I’ve focused on:

• Street Fighter IV: 2009-2011

• Marvel vs. Capcom 3: 2011-2016

• Guilty Gear Xrd: 2014-2019

• Dragonball FighterZ: 2018-2023

• Guilty Gear Accent Core Plus R: 2020-2021

• Vampire Savior: 2022-Present

There is some overlap, and I never truly stopped playing most of these games (I will never turn down a chance to get some MvC3 matches in), but at most, I think I focused on two games at the same time, and that was usually after I’d spent enough time on the earlier game that I had the mental bandwidth to try something new. For example, when I started playing Xrd in 2014, I’d already been grinding MvC3 for ~3 years at that point, so I could stay sharp at MvC3 without it being my full-time job.

Another caveat is I know there are people who are exceptions to this—folks who picked up their first fighting game, played the hell out of it, and managed to get top 8 in a major having only really played that one game. I still those people are the exception rather than the rule.

Why is Being T-Shaped Helpful?

I don’t focus too much on the “why” in the video, aside from surface-level stuff like how different fighting games emphasize and teach different skills.

I do think one important component is learning how to learn. When you pick up your first fighting game, you have no idea what to look for. What makes a move good? What is frame data? What are some standard counterplay options for divekick moves? You’re not going to have easy answers to these questions (because you’re new!), so you have to (slowly) figure things out on your own as you stumble through the game. When you pick up your next game, you’ve already pre-solved a bunch of these problems. You’ll see a divekick and already have two or three options in your head to deal with it. You’ll go through the fighting game learning process much more quickly now that you have a foundation.

A Final Random Aside

I felt like I needed footage of me playing Blanka in Street Fighter IV for this video, and I had exactly one source of footage: the very first YouTube video I ever uploaded. It’s on my old channel, and I uploaded it back in 2010 (I think I am officially a senior citizen now).

I was at a friend’s apartment, and we were taking turns playing online matches against random people online. We ran into Sako, and somehow I managed to get a win, after which he promptly booted us from his lobby. I thought it was funny, so I uploaded it to YouTube with the respectful title of “Sako? More like Sucko! America Redeemed.” Shockingly, everybody thought I was a jerk (it was a joke title, I swear!) Anyway, if you’d like to watch the match in its entirety, here you go:

I believe fighting games can be bigger than they currently are. I made a whole video about the idea. The elevator pitch was that fighting game developers had limited the reach of their genre through a decades-long series of unforced errors. I’m pretty convinced that if someone could make a fighting game that avoided those missteps, it could reach an audience an order of magnitude larger than any fighting game to date. I think some of the key boxes to tick are:

1. Rollback netcode. Good rollback netcode is what makes online play for fighting games function properly. Only very recently did this become the standard (games without rollback are rare these days).

2. Good UI/UX. Even though good implementations of rollback netcode are common now, fighting game devs continue to drop the ball on their user interface/user experience. Guilty Gear Strive’s terrible tower lobby system is the most obvious example of horrific UI/UX in a modern game, and sadly, it’s not alone.

3. Free to play (or some reasonable business model). I think something like Dota 2’s model would work well in a fighting game context (where all gameplay-related things are free and the only things costing actual money are purely cosmetic). I think you could also do something like the original Overwatch, where you do pay a fixed price for the game, but then you have access to new characters, maps, etc., as they’re released, and again, the only reason to spend money is on cosmetics.

4. One button special moves. I did a whole video about this as well, but the short version is that while I don’t think traditional special move inputs are difficult, a lot of people perceive them as difficult, so easier inputs can help folks get past that mental barrier. One button specials also make a game more easily playable on a variety of input devices, which widens your potential audience.

Points #1 and #2 are pretty widely acknowledged, but I think it’s important not to underestimate the impact of points #3 and #4 and how they work together. If your game is free to play, and you can play it properly with a standard keyboard/mouse, anybody with a computer is a potential player (at least as long as your game can run decently well on lower end systems).

Fighting game devs have finally come around on rollback, so the unforced error of poor netcode is (hopefully) a thing of the past. Good UI/UX is still elusive. Some devs seem to struggle with it (Arc System Works) while others are better (I think Capcom did a solid job with Street Fighter 6). However, traditional fighting game developers have been very reluctant to seriously consider either #3 or #4, despite the upside potential.

Riot’s 2XKO (formerly known as Project L) has been in the works for years at this point, and within the last year or two, they’ve started bringing it to fighting game tournaments for people to try out. As I type this, they’re running a closed beta test they’re calling Alpha Lab. I managed to get into Alpha Lab, so I've been playing the game over the past few days.

Ever since the game was announced, I was 99% certain they’d hit all four points I listed above. While some smaller indie titles (like Fantasy Strike) had also checked those boxes, 2XKO is the first one with both the IP strength and the budget to really test my hypothesis. Anyway, I think I’ve buried the lede enough—while I still firmly believe in the theoretical potential reach of traditional fighting games, 2XKO is not the game that’s going to realize it.

Hitting all four points is necessary but not sufficient. While extremely important, those four things are ultimately features that are attached to a game, and people have to want to play that game for any of that to matter. I’ve been extremely impressed that Street Fighter 6 has consistently had a 24-hour peak player count of ~25,000—that’s a massive success! I think plenty of people will play and enjoy 2XKO, but after sitting down with the game, I’m having a very hard time imagining it even sustaining SF6 numbers once it’s been out for a bit and the dust has settled.

2XKO’s Controls and “Easy” Special Moves

Here’s what 2XKO’s control scheme looks like (for now, at least):

There’s a lot going on here!

As expected, 2XKO has one button special moves, so all specials are performed by hitting a button (the Special 1 or Special 2 button), or one of those special move buttons plus a direction. Outside of special moves, you have light, medium, and heavy attack buttons, and in many cases, you do different moves depending on whether you press just the attack button or if you press the attack button plus a direction. 2XKO also allows you to call the character who’s tagged out to have them come in for an assist. This is performed by pressing either backward or forward plus the tag button, which lets you pick between their two available assists.

So you've got a lot of normals/specials that require a button plus a direction, and those inputs exist alongside assists that require that you press the tag button plus a specific direction. This makes it so you usually can't do a move while calling an assist simultaneously. This is something you do all the time in assist-based games, and it’s been driving me nuts for the entire beta.

Let’s look at an example (I’m going to use numpad notation): I'm playing Illaoi on point with Ahri tagged out. I'm doing a block string and decide I want to do 6+S2 with Illaoi on point while calling Ahri's back assist (4+T). Obviously it's impossible to press both left and right at the same time, so I need to do one of the two things first, and then wait a few frames and do the input for the other. I'd way rather just have two tag buttons with each one dedicated to one of the two available assists. They took great pains to avoid difficult inputs, but their solution makes many common things you’d like to do maddeningly difficult.

Another common downside of one button specials is that you end up needing more buttons than you would with traditional inputs. At minimum, you need six buttons to play 2XKO: light attack, medium attack, heavy attack, special 1, special 2, and the tag button. You can dash by pressing L+M at the same time, but it’s much easier to dash consistently with a dedicated dash button, so that’s a seventh button. You can parry by pressing L+H, but again, there are tangible benefits to having a dedicated parry macro, so that’s button number eight.

You have to use all of these buttons all of the time, so it becomes a lot to juggle. Contrast this with Marvel vs. Capcom 3, another assist-based game, where if you’re playing on stick or leverless, your index finger, middle finger, ring finger, and thumb can all comfortably rest on the four main action buttons (L, M, H, and S).

It’s not just me that feels this way—I’ve been seeing similar impressions all over social media:

I can see this being less of an issue if you play on pad specifically, but that seems less than ideal. One of the big benefits of simplified inputs is that they’re supposed to make your game more input device agnostic, so it seems strange to feel like I need to play on pad to be comfortable.

The System Mechanics of 2XKO are Complicated as Hell

If you follow that Alioune thread, he goes on to say:

I pretty much agree 100% with what he’s saying. Why is there both retreating guard and push block? Why does crouch block let you block mid attacks, but low parry doesn’t work against mid attacks? Why do you need to juggle all of these defensive mechanics to escape the game’s many oppressive pressure sequences? Other assist-based fighters tend to focus on blocking plus one main defensive mechanic that offense/defense revolves around (reflect in Dragon Ball FighterZ, push block in Marvel vs. Capcom 3, etc.)

In many ways, 2XKO is structured more like Marvel vs. Capcom: Infinite or BlazBlue: Cross Tag Battle, where you play a team of two characters and can switch between them in a bunch of different ways. My experience in the beta so far is that the game can be immensely visually confusing. After calling an assist, you can switch to playing the assist character by tapping the tag button, so anytime an assist is called, you have to consider whether a character switch is coming. Something about how the camera focus changes between the characters during a tag adds to the confusion. I find it a much harder game to follow than more standard assist-based fighting games like DBFZ or Marvel vs. Capcom 2 or 3.

It’s also really easy to accidentally tag in your assist character. For example, let’s say I’m blocking a move and I want to call my assist right after I exit block stun. I press back plus tag to get my back assist, but then I realize I may have pressed it a bit early (while I was still blocking). You can’t call an assist while blocking, so I press back plus tag again. But then it turns out I actually didn’t press the first attempt too early, so that first assist call worked, and now my second attempt to call the assist has become me accidentally tagging in my assist character.

Duo Mode is Weird

Most fighting games are a single player vs. another single player. This has its pros and cons. You’d expect it to make online matchmaking easier than in team-based games, since you just have to match up two individual players (rather than trying to assemble and balance teams of multiple players of different skill levels). 

Some aspects of 1v1 can be taken as a pro or a con depending on your personal taste; whether you win or lose is 100% on you. I personally like this, but team-based games give you someone else to blame when you lose, and sometimes your team can carry you to victory, so you’ll get to see some wins regardless of your individual skill.

The biggest online games are team-based (let’s set aside battle royale games for now, since they’re kind of their own thing), so it makes some sense that fighting game devs continue to chase the dragon of team-based modes, hoping to unlock a larger potential audience for their game. 2XKO has duo mode, where you can play with another human player controlling the other character (rather than you just controlling all of your characters like in a traditional assist-based game). I’m skeptical that this is going to expand the audience of the game.

Whether a game is 1v1 or team-based, the vast majority of players will solo queue when playing online. In a 1v1 game, that means you’re playing games against opponents the matchmaking system finds for you (as opposed to arranging a match against a specific player beforehand and meeting up with them). In a team-based game, this means queuing up and letting the game find you a team to play with (and a team to play against, of course). 

How well this works in team-based games depends a lot on the design of the game itself. When you solo queue and just play on a team of random people, you’re almost certainly not going to be communicating with them. Most people won’t use voice chat. At best, people might type some random stuff in the chat (if you’re playing on PC, at least). This means that the team-based games best suited to solo queue are the ones where without any use of spoken or written language, you can coordinate and play with your team somewhat effectively.

Duo mode seems like it really requires you to play with a teammate who you specifically regularly practice with, and then you’d only play duos with that specific person, and you’d use voice chat to coordinate during the match. I have a strong feeling that the number of players who do this will be miniscule. 

I’ve seen people push back against criticism of the game by saying that playing duos solves whatever problem you’re having. I think this rings hollow because this is a mode most players won’t meaningfully engage with. If duos is the one true “correct” way to play the game, I think it’s going to be dead on arrival.

What I’d Want

I don’t expect this to happen, but I would be way more comfortable playing the game if there was an optional control scheme that was something like:

• Light attack

• Medium attack

• Heavy attack

• Assist 1

• Assist 2

Many of the inputs would stay the same (L+M could still be dash, M+H could still be throw, etc.) The main changes would be:

1. Specials would all be quarter circles plus an attack button. This would mean on a stick or leverless controller, you would have access to all of your normal and special moves without having to shift your hand position at all. Supers could be a quarter circle plus two attack buttons.

2. Assists would be done through pressing either the Assist 1 or Assist 2 buttons. This would mean you would not need any directional input to call an assist, and it would mean you could be doing whatever directional input you want while calling whichever assist you want. For example, you could walk forward while calling your back assist, or you could block while calling your forward assist.

3. Tagging in your other character (either through quick tag and handshake tag) could just be done by pressing the two assist buttons together. This way, you’d avoid accidentally tagging in your assist character like I described in my example earlier. Dynamic save (burst) and tag launcher could also be done with this same input.

I don’t realistically expect them to offer this as an option, but it’d be a lot more comfortable/intuitive (for me, at least). As it stands, I’m not sure who the game is really for. Their choice to go with one button specials makes it seem like they’re trying to get new players into the genre, but everything else about the design of the game seems 100% focused on satisfying existing fighting game enthusiasts. They’re sending mixed messages.

Final Thoughts

When any new fighting game comes out that I’m interested in, I know that the odds of me sticking with it for years and years is pretty low. I already have fighting games I love, and they aren’t going anywhere, so it takes something special to get me to permanently reallocate my fighting game time and play something new. 

Even if 2XKO is not a game that I end up sticking with, I’m still rooting for its success. I think a lot of older school fighting game devs will be watching closely. A successful 2XKO sends the message that free to play is a viable business model for fighting games. It sends a message that being transparent and communicative with your community matters. It sends a message that quality of life features like not having another player’s bad internet connection ruin my experience are worth implementing. I think if 2XKO doesn’t gain much traction, other devs will not learn these lessons and just stick to what they know and continue to play it safe.

Who knows though—while my gut feeling is that 2XKO is not going to light the world on fire, I could be wrong. I’ve been wrong about plenty of stuff before, so we’ll have to wait and see what happens.

As I write this up before releasing the video, I have no idea if the video will go over well, or if it will crash and burn pretty damn hard. Go watch it before reading on! This writeup is mostly going to be a postmortem where I expand on some of its ideas and hopefully clear up some possible misunderstandings.

I feel like this video will probably make a few people mad who prefer traditional motion inputs. For the past two years or so, I haven’t really been playing any modern fighting games at all. I’ve instead been playing Vampire Savior nonstop—a game from 1997 with a ton of weird inputs and a lot of design sensibilities that you don’t see in recent games. I love it.

At this point, I honestly don’t know if I’ll ever seriously play a modern fighting game again. This doesn’t really have anything to do with traditional vs. simplified special move inputs—it’s mostly about the direction the genre has taken in recent years (a topic for another day). The point is that while I still pay attention to recent trends and make these kinds of video essays (it’s fun!), I'm having a great time playing old-timey games, and that’s probably going to continue for the foreseeable future. The video is not me trying to say what should be happening. I’m just trying to explore what is happening (and why).

I absolutely stand behind the why of the video—I really do think the two primary driving forces behind simplified inputs are perceived accessibility and making games controller-agnostic. The word perceived plays an important role here. I allude to it in the video when I talk about Smash being weirdly hard to control despite it having the reputation for being approachable. Making inputs simpler mostly matters because people believe it matters (and not necessarily because it actually does). The controller-agnostic thing is more based in dollars and cents. If your game plays fine on a computer keyboard and you make it free to play, your potential audience is massive.

One thing I didn’t have room to address was that I believe this trend is going to force certain types of design decisions and push the genre in a certain direction. I pointed to Street Fighter 6’s modern controls being an issue because one button fast invincible moves are problematic in the context of Street Fighter specifically. Street Fighter neutral is deliberate and slower paced (compared with most other fighting games). Balancing the startup and invulnerability frames to make reversals strong (but not too strong) is tough in this type of game. 

However, this sort of thing is way less of an issue in some other fighting games. In Ultimate Marvel vs. Capcom 3, Spencer’s bionic arm super (I know it’s called bionic lancer, go away) is invulnerable, has just 7 frames of startup, and the input a simple quarter circle back. It has a massive hitbox that hits all the way up near the peak of normal jump height, and it goes nearly full screen. Depending on available resources, if you hit somebody with bionic arm, you can kill the character. You can even use the team hyper combo mechanic to do bionic arm by just pressing your two assist buttons at the same time, ignoring the normal motion input and essentially making it a one button move. Despite all of that, bionic arm is pretty well balanced, and Spencer is generally considered an upper mid tier character at best. How is that possible? It’s a combination of factors:

• Moving your character around in marvel is pretty safe. If you dash on the ground, and Spencer does bionic arm, you can just hold up back to cancel the dash into a jump to block it (and punish).

• Lots of stuff is more invincible than bionic arm. These other moves aren’t as quite as quick to startup, but you can easily react to the super flash and beat bionic arm if you’re in position and have the resources to answer back with your own invulnerable move.

• In an assist-based game like marvel, you protect your assists and your assists protect you. You can confirm that Spencer is blocking your assist before approaching to keep yourself safe from bionic arm.

• Most importantly, marvel is an extremely fast paced game. You can’t just sit there and look for stuff to react to. Bionic arm is also only available on the ground, so Spencer committing to staying on the ground is a death sentence. Aerial combat is really important in marvel and you’re going to spend a lot of the match airborne.

If you put bionic arm in Street Fighter, it’d be busted as hell, but the core design of marvel makes bionic arm a pretty reasonable move. I think this is a big reason why 2XKO ended up being a marvel-inspired tag fighter—if you’re going to commit one button special moves, emulating a game where you already know they work makes a lot of sense. It’s kind of interesting or paradoxical or something that one-button specials are forcing them to make a game like marvel (a series not known for being super approachable).

I don’t know that this mini essay has any kind of grand thesis that I’m working toward, and I’m out of things to say for now. Come play Vampire Savior with me sometime!

I’m not currently planning on making Think, Don’t Mash! videos specifically for Street Fighter 6. While I used Street Fighter V to explain the concept, the series was never meant to specifically be about Street Fighter. I was trying to show how the core idea of starting with a small number of key moves and then slowly expanding upon that over time can be applied to any fighting game.

That being said, some folks have asked for my take on Street Fighter 6. I responded to a few YouTube comments with how I approached SF6, so I thought it might be helpful to put that information into a blog post and expand on it a bit.

Disclaimer: I need to come clean. I played Street Fighter 6 for around a month when it launched and then went back to focusing on other games. This is one of the reasons I’m not really doing Street Fighter 6-focused stuff—it seems like it’d be disingenuous. 

In my month of playing Street Fighter 6, I started with the same plan as I used in Street Fighter V and then expanded on it a bit. For a quick refresher, the Street Fighter V plan was to begin playing a character using only:

• Universal grounded movement options (walking and dashing backward and forward)

• A good grounded poke

• My normal throw

• An anti-air

• An anti-fireball tool

I realized pretty quickly this plan needed a few modifications to work properly in Street Fighter 6. We’ll still need the same grounded movement options, we’ll still use normal throws, and we still need an anti-air—these three are staying the same. The others need to be tweaked or expanded upon.

Picking a Grounded Poke

Picking a grounded poke was always my starting point with a new character in Street Fighter V. I just looked for a poke that had good reach and was safe on block. I picked Chun-li’s standing heavy punch in the original Think, Don’t Mash! video. Chun’s standing heavy punch  made it into SF6 more or less intact, so I should be able to use it the same way here, right?

Street Fighter V has the V-System, where you have a gauge that fills up during the round (mostly due to you taking damage), and you can spend it in a few different ways. The most obvious way is to activate your character’s V-Trigger once the bar is full, which temporarily either powers up your character or gives them access to new tools. At lower levels of play, I don’t think V-Trigger matters too much—if the player remembers to activate it at all, they aren't going to use it in a way that will drastically change the flow or outcome of the match. Because the V-Gauge is always empty at round start, it’s not a factor until well into the round.

Street Fighter 6 swaps out the V-System with the Drive System. It has some similar uses to the V-System, and many of these don’t matter much at lower levels of play, so you can safely ignore them starting out. However, there is one specific use which you will see A TON OF at lower levels of play: Drive Impact. 

You perform drive Impact by pressing heavy punch and heavy kick at the same time (or you can just assign a button to it). It’s an attack that has two hits of armor (meaning it can blow through up to two hits without being stopped). It starts up pretty quickly. If it hits as a punish (which it usually will if it armors through an attack), it crumples the opponent and lets you hit them with a combo for free. Also, unlike SFV’s V-Gauge (which is empty at the start of a round),  SF6’s Drive Gauge is full at round start, meaning from moment one, Drive Impact is a threat (and it only costs one chunk of the gauge, so you can use it six times in a row before running out). 

If you try to use Chun’s standing heavy punch to control the ground the same way we did in SFV, a random Drive Impact can blow through it and destroy you.

Dealing with Drive Impact

Plenty of games have moves that can beat almost anything. Ryu can do an invincible dragon punch to beat any grounded button you press. Ryu can’t just dragon punch all day though because if he misses, he’s wide open—you can hit him with a full combo for free. Drive Impact does not work like this. If you block it and you’re mid-screen, it pushes you back but the attacker is completely safe. If you try to block it and you’re close enough to the corner, it causes a wall splat, letting them get a free combo on you. Blocking is not the answer.

Luckily, the solution is simple: you can do your own Drive Impact to beat theirs. The second Drive Impact will always win, so if you’re just sitting there blocking and see your opponent use Drive Impact, you can Drive Impact them right back, giving you the chance to hit them with a full combo.

So what stops the match from devolving into a game of chicken, with both players sitting and doing nothing, waiting for the opposing player to drive impact so they can beat it with their own? In Street Fighter, some normals can be canceled into special moves. Any normal that is special cancellable can be canceled into Drive Impact. This means that if you poke with a special-cancellable normal and your opponent Drive Impacts through it, you can always cancel your poke into your own Drive Impact and win:

There are two important take-home messages:

1. You need a poke that’s special-cancellable. When you’re just starting out, if you’re going to use just one poke, it needs to be special-cancellable.

2. You need Drive Impact in your own arsenal from day 1.

Instead of Chun’s standing heavy punch, we could instead pick back/forward medium punch (both back and forward give you the same move). It’s a good poke that’s special-cancellable, so you can keep yourself safe from random Drive Impacts in neutral.

Anti-fireball Tools

The most universal anti-fireball tool in Street Fighter has been a predictive forward jump over the fireball (unless you have some wildly floaty jump like Dhalsim). If you predict a fireball correctly, you’ll get your best jump in combo. This is still true in Street Fighter 6. There have also been character-specific tools to do things like slide under fireballs. These also still exist in SF6. When first starting out, these are still both solid options to pick to deal with fireballs. SF6 has also added a new defensive option which is great for this purpose: Drive Parry.

Drive Parry is done by pressing medium punch plus medium kick (or you can assign it to a single button, just like Drive Impact). Drive Parry mostly functions the same as blocking. The difference is that if you block a special move (like a fireball), you lose some of your own Drive Gauge (instead of taking chip damage like in most fighting games). But if you Drive Parry the special move, you don’t lose any Drive Gauge—you basically break even, since activating Drive Parry costs half a chunk of Drive Gauge, but parrying a projectile rewards you by giving you half a chunk of Drive Gauge back. This means you can just hang out and defend against fireballs for no real cost or risk. If you have the life lead, your opponent can’t just sit on the opposite side of the screen and chuck plasma—they have to come to you.

Putting the Pieces Together

Putting this all together, the SF6 version of the Think, Don’t Mash! starting plan would be something like:

• Universal grounded movement options (walking and dashing backward and forward)

• My normal throw

• An anti-air

• A good grounded poke that’s special cancellable

• Drive Impact

• An anti-fireball tool (probably Drive Parry, even if you’re still going to sometimes predictively jump over fireballs too)

*New stuff is bolded

For Chun, I’d try starting with back/forward medium punch as my special-cancellable poke, and I’d try standing medium kick as my anti-air. You can still use standing heavy punch sometimes too, but it’s good to have a different special-cancellable poke to fall back on if they’re relying too heavily on Drive Impact to win neutral.

I think Street Fighter 6 is a harder game than Street Fighter V to play starting out, just because at even the most basic level, you need to use more tools. The main high risk high reward option you needed to shut down pre-SF6 were jumps. Jumps in Street Fighter usually take around 45 frames (three-fourths of a second). This gives you a lot of time to react and ant-air, since human reaction time is somewhere a bit under 20 frames. Of course, even pros miss anti-airs sometimes because of the mental stack—when you have to look for lots of different things and react differently to each of them, it becomes harder to react quickly to any one thing. Drive Impact has just 26 frames of startup. This is still reactable, but it’s almost twice as fast as a jump, and jumps are already something new players will struggle against at first. You have to be ready with two separate actions to deal with the two universal high risk high reward options that exist in the game. This also brings us back to the mental stack—it’s even harder to anti-air consistently in SF6 relative to SFV because you need to devote some portion of your brain to reacting to Drive Impact as well.

Anyway, that’s all I’ve got for now. Hopefully this at least gives folks a decent starting point.

I bought a Cross|Up™ (I’m just going to call it a cross-up for the rest of this article, sorry Hit Box Arcade™). I thought it’d be interesting to discuss it a bit!

tl;dr:

If you want to skip to the end:

• The cross-up is straight-up better than a standard arcade stick, so if you’re going to buy a stick with a lever in it, you should get something with a cross-up style layout (assuming it’s available and within your budget).

• Completely leverless controllers like the Hit Box are still probably objectively the best overall fighting game controllers.

• How good a given controller “objectively” is only matters at the margin. No controller will magically make you go Super Saiyan, so just play on what you like best.

• The build quality/extra features on the cross-up are really nice.

How It Works

If you’re not familiar with it, you might be wondering what the hell all of those extra buttons are for. The lever is your left analog stick and the black buttons are the same attack buttons you’d find on any regular stick. If you ignore the buttons with the white rims, you could use this exactly like you’d use a normal stick. The buttons with white rims are intended to be used as the dpad, giving you another way to input directions. By default, it’s configured like this:

This allows you to do certain kinds of directional inputs more easily than you could do them on a lever alone. For example, in Guilty Gear Strive, a lot of supers have an input that’s 632146 and then an attack button. On a lever, it kind of feels like three distinct steps—you do the half circle, and then swing the lever back to forward, and finally hit the attack button:

For a lot of characters, this is the one invincible move they have, so you might want to use it as a reversal, but consistently getting the input quickly and accurately in reversal situations was inconsistent for me. On the cross-up, you can divide the motion between your two hands. My left hand can do the half circle part, and my right hand can hit forward and the attack button. This makes it feel more like just two steps:

In practice, this worked exactly like I’d hoped—it’s way easier/faster vs. using the lever alone.

With the extra directional buttons, it seems like I have a few possible paths with the cross-up:

1. I don’t end up using the extra buttons really, so it’s “just” a stick

2. I use the buttons for certain specific use cases (like my Guilty Gear Strive example above), but 95% of the time, I ignore them

3. I use the buttons a lot and they’re integrated into my moment-to-moment play

I’ve had the cross-up for about six months now, and path 2 seems like the most likely one. Using the directional buttons requires you to (slightly) reposition your hand, so l go to them for specific use cases, but most of the time, my fingers rest on the main action buttons. This is the reason that I don’t think a cross-up can be as objectively good as a completely leverless controller like a hit box. 

On hit box, your fingers are always resting on the directional buttons, so there’s minimal hand repositioning, and aside from certain types of inputs (like half-circles), you get the advantages of a cross-up all the time (and not just when you decide to reposition your hand).

Also, while the two input methods for motions do allow the cross-up to do certain things better than a hit box, there’s nothing to stop stop you from combining the hit box and cross-up ideas into the Infinity Gauntlet of fighting game controllers.

The Software

You can use the Cross|Up Designer software to change the inputs associated with each of the different buttons, and you can also change things like the SOCD method used. There’s a rocker switch on the cross-up itself that allows you to switch between three different main profiles (and you can have even more with a bit of extra work).

For example, I have this as one of my main three layouts for when I play Vampire Savior:

I swapped the dpad up and down buttons. With the up button next to the left/right buttons, it's really easy to hit 9 (up+forward) with the buttons. This allows for crazy fast instant air dashes by pressing up+forward with the movement buttons and then tapping forward on the lever as soon as I release the buttons. I can do faster IADs than even double tapping up+forward on a hitbox.

I haven’t messed around too much with the software (the rabbit hole goes deep!), but it really opens up your options and allows the cross-up to be used for nearly any game (especially with the option to plug a Wii nunchuck into it).

Final Verdict

If you saw the tl;dr at the top, you already know what I’m going to say here. If you’re planning on buying a stick with a lever, just buy a cross-up or a cross-up equivalent (assuming it’s within your budget).

The extra directional buttons are great to have. Whenever there’s a specific input that’s giving you trouble on the lever alone, the buttons offer ways to work around it and troubleshoot.

Watch the video first and then come back here:

This video started off with a pretty different concept from the final product. Originally, it wasn't going to be centered around Justin Wong at all. The title was going to be something like How 95% of Marvel Players Lost to Morrigan at Character Select. The parts about why the  Morrigan/Doom/Vergil team is so good and the five main strategies to beat it would still be the same. But instead of using Justin Wong as a positive example of how to use the strategies and win, I was going to look at some of the players who struggled against Morrigan and analyze why.

I don't think this would've gone over very well.

People would (rightly!) point out the fact that I'm a nobody who hasn't won anything, so who am I to judge. I don't think this would discredit my points/analysis or anything, but I have a tough time being that antagonistic, even though the internet seems to like that sort of thing. I think the spiciest title I considered was Why RayRay Almost Never Beat Chris G.

Luckily, nobody reads these posts, so I can talk about it a bit here. If you've already watched the video, you'll know that I lay out five strategies to deal with Morrigan/Doom/Vergil. I allude to it in the video, but I'll state it plainly:

Morrigan/Doom/Vergil is an end-game team and possibly the best team in the game. It is not possible to guarantee a win by picking a specific team to beat it. However, it is 100% possible to lose to Morrigan/Doom/Vergil at character select.

I spent the better part of a decade playing and spectating Marvel 3, and it always blew me away how a bunch of excellent players stuck to their guns and played their main team into Morrigan/Doom/Vergil, even though it was garbage against it. I wasn't the only one to notice this!

Since I already threw RayRay under the bus, let's look at his main team as an example. RayRay mains Magneto/Doom/Sentinel—a great team. Playing that team, RayRay made top 8 at multiple majors, including EVO. That being said, I think it struggles against Morrigan/Doom/Vergil. Again, my five strategies vs. Morrigan/Doom/Vergil are:

1. Maul: Pressure hard and try to land a hit during the opening gambit. If you can't, keep up the pressure and find a hit quickly.

2. Dark anchor: Use a strong anchor character you can exploit meter and level three x-factor to reverse OCV.

3. Stall: Evade fireballs until astral vision runs out to deny Morrigan another activation.

4. Fight Straight Up: What it says on the tin.

5. DHC Momentum: Spend two bars to DHC between two characters in a way that lets you pressure Morrigan.

What can RayRay's team do? Magneto can stall pretty well. He can use flight and plink dashing to avoid fireballs and missiles. He can throw some Magnetic blasts to negate a few fireballs. Once astral is over, he can fight back against Morrigan. He can use his beam to punch through fireballs and snipe Morrigan/Doom on the ground. Once there aren't astral fireballs coming from behind him, he can call Doom beam assist to gain some momentum and try to move in.

The problem is that there is no backup plan. If Magneto is about to die, RayRay can DHC from Magneto to Doom, but this is just going to leave Doom at full screen with no meter. At this point, Doom can stall out astral, but this doesn't really get him anywhere. There's no way for Doom to ever re-establish momentum and bring the fight to Morrigan.

If Magneto just dies early, RayRay will likely have two bars for Doom/Sentinel, but there's no way to really spend that meter and bring the fight to Morrigan. RayRay basically has to win the point war, stalling out the first astral activation, playing clean, and then landing a hit on Morrigan and killing her. This doesn't guarantee victory (Doom/Vergil is still a very real threat), but the opposite scenario, where Mags goes down early, very nearly guarantees defeat.

We can contrast that with Filipino Champ's Magneto/Dormammu/Doom team. As another point Magneto team, at match start, it can do the same thing RayRay's team does (stall astral and fight Morrigan straight up), but then it has some very important additional options. If astral isn't active, Magneto can usually find a DHC from magnetic shockwave into Dormammu's stalking flare. This will clear the screen and force Morrigan to block the stalking flare. With that momentum, Dormammu can fight Morrigan quite well. He can keep forcing her to block pillars while calling assists to keep control of the screen.

If Magneto dies on the Magneto/Doom/Sentinel team, Doom/Sentinel is almost a lost cause against Morrigan. If Magneto dies on the Magneto/Dormammu/Doom team, Dormammu/Doom still has a pretty real chance against Morrigan/Doom/Vergil. Chris G ran into both players at EVO 2014, so you can watch for yourself:

I don't really know how to end this post. I guess that up through Marvel 3's original EVO run (ending at EVO 2017), this story stayed the same—lots of players losing to Morrigan/Doom/Vergil at character select. I hope this doesn't happen at EVO 2023.

Understanding the Street Fighter Series as Capcom's Reaction to Criticism

For the last two decades, each Street Fighter game has felt like a reaction to the criticisms of the preceding game. This is not a bad thing! I'd much rather have a developer be responsive and try to improve things for each new iteration of a series. I think this is a useful framing though—when trying to understand where Street Fighter is going, it's important to understand the path the series has taken.

In Street Fighter III, people liked the idea of parries but were put off by the execution needed to use them effectively. Parries also largely eliminated zoning from the game, which feels weird given the Street Fighter series invented the fireball. People also didn’t like how Capcom largely abandoned legacy characters, only bringing back Ryu, Ken, Chun-li, and Akuma.

Street Fighter IV tried to hit each of these things like they were working through a bulleted list. The focus attack mechanic was intended to be a more user-friendly parry that didn’t invalidate zoning. They also leaned hard into legacy characters—at launch, the arcade version of Street Fighter IV had the entire cast of Street Fighter II with just a handful of newcomers alongside them.

SFIV was well-liked and played an important role in growing the fighting game community, but plenty of people had issues with the game. For most of the game’s lifespan, set play was very strong—knock downs would give the attacker an extremely strong mix-up situation, and if successful, the mix-up would lead to another knockdown and it would loop into itself. This escalated as people discovered unblockables (which Capcom tried to address through adding delayed wakeup).

While set play was strong, SFIV could feel like an overly defensive game in some situations. The crouch tech OS really weakened normal fighting game offense. The game also had no input buffer, and you had characters like Sakura who were pointless to play unless you could consistently hit one frame links. For a game that was trying to be more accessible, this felt out of place.

Street Fighter V is easy to understand through this lens of reacting to the previous game. SFIV had zoning and lots of people don’t like zoning, so characters like M.Bison and Nash in SFV have tools to negate projectiles completely. Capcom added an input buffer so that there were no one frame links. They got rid of the crouch tech OS so standard strike/throw offense could exist again. They also had multiple wake-up timings for the defender to use to weaken set play.

Of course, SFV had plenty of criticisms—for the first chunk of the game’s life, the characters and their game plans were very homogenous, and for the most part, their game plans were pretty simple and straightforward. This got a lot better with later seasons of the game (especially after Yoshinori Ono left Capcom). 

Most of the SFV’s issues weren’t related to the actual mechanics of the game. At launch, it was missing lots of standard fighting game features and modes. The game had terrible input delay and has ongoing netcode issues. To their credit, Capcom stuck with the game and addressed nearly every issue (with the exception of the netcode), but the game’s rocky start really set back the growth of Street Fighter as a competitive staple in the fighting game community.

Street Fighter 6—The Next Step on the Path

Street Fighter V launched as a $60 retail game, but it was a terrible value proposition. I really do think it was intended to be free to play, but the idea was scrapped at the last minute, so Capcom ended up charging full price for something that didn't justify it. This presented kind of a fork in the road for Street Fighter 6. SF6 could either actually be free to play, or it could be really full-featured at launch to justify its retail price tag. Capcom opted for the latter, which I talked about in a previous post.

I think the big question at this point is where SF6 is headed mechanically. Because people’s criticisms of SFV were largely about things outside of the mechanics of the game itself, SF6 feels like the first Street Fighter title in a long time where Capcom has had a blank slate, without a strong need to design around complaints about the previous title. They’ve taken a kitchen-sink approach and tried to include mechanics from a bunch of previous Street Fighter titles, and it's difficult to predict where the game will end up as a result of this.

The Drive System

The drive system is a bunch of different ways you use and interact with the drive gauge at the top of the screen (right beneath the health bar). It has six chunks, and you start each round with a full drive gauge, so you immediately have access to the whole toolbox. If you run out of drive gauge, you go into burnout, where you lose access to all the drive system mechanics, take chip damage, and attacks against you cause additional blockstun. Let's look at the different ways you can spend this meter.

Drive Impact

Input: HP+HK

Cost: 1

Compare with: Focus Attack (Street Fighter IV)

You do an attack that gets two hits of armor from the first frame you activate it, so it will blow through lots of attacks. On hit or block in the corner, it can lead to good damage since it causes a wall splat. If it hits as a punish, it causes a crumple and you get a combo mid-screen also.

If your opponent is going crazy using this, there are obvious counters. You can do your own drive impact (the second drive impact will always win since its armor will last longer). You can also neutral jump and then punish with a jumping normal as you fall after the drive impact has whiffed. At lower levels, people probably aren’t going to react and punish it consistently though, so this mechanic seems like it’s shaping up to be The Scrub Killer™.

Drive Impact makes certain “standard” Street Fighter stuff like Ryu cr.MK xx hadouken much weaker (since you can drive impact between the cr.MK and the hadouken). Even at higher levels of play, it may remain important, beating certain pokes (on reaction or prediction), invalidating certain specials (being able to Drive Impact through on reaction), etc. It also becomes scarier against an opponent in burnout, since they can no longer counter with drive gauge mechanics like their own drive impact or parry.

Drive Parry

Input: MP+MK

Cost: 0.5 (more if you hold the buttons)

Compare with: Parry? (Street Fighter III)

Parry is kind of two separate things that each need to be considered on their own. 

Regular parry: generally, parry will not affect frame data when compared with blocking. If Chun-li’s st.MP is +1 on block, it’s still +1 on parry. What parry does do is cover all possible blocking options. With parry, you don’t have to guess whether a move is a cross-up or whether it hits high or low. You also can block during parry recovery, so you can never be punished for attempting a parry unless your opponent throws you (it’s likely pretty low risk against characters without command grabs).

If you time your parry well so that it parries something in the first two frames of activation, you get a perfect parry. Perfect parry has just one frame of recovery, so you can punish tons of stuff. When you perfect parry, your opponent also can’t cancel their attack into anything, so your punish is guaranteed.

Because perfect parry has just a two frame window, I think it will take time for even high level players to begin to consistently take advantage of it, but I think it could end up eventually being game-defining. For example, if you can perfect parry on wakeup (you have to time it properly, so this isn’t easy), you can punish a lot of meaties. Similarly, if you’ve blocked something where your opponent is +3, you can attempt to perfect parry them trying to take their turn with another quick attack with their plus frames. Perfect parry has the potential to really disrupt how we think of turns in a Street Fighter game.

Drive Rush

Input: →→

Cost: 1 (from a parry) / 3 (from a normal)

Compare with: FADC (Street Fighter IV)

Drive rush lets you cancel a parry or a special-cancellable normal into a fast and far-moving forward dash. If you do a normal out of drive rush, it gains additional frame advantage. In the SF6 beta, nearly every normal is negative on block, so drive rush seems key to prolonging pressure, though this is balanced by its three bar cost.

Parry into drive rush costs a lot less meter though, so you can do parry (whiff) into drive rush for 1.5 bars. This means every character in the game can spend 1.5 bars to have a pretty crazy mobility option. There is also some slowdown during drive rush, so you can do this to have more time to visually assess a situation and react to stuff (similar to how you could use YRC in the Guilty Gear Xrd series).

Overdrive

Input: Your normal special move input + one additional punch or kick

Cost: 2

Compare with: EX moves (Street Fighter EX/III/IV/V)

Street Fighter games have had EX moves for a while now, where you could do an enhanced version of a special move. For example, EX fireballs usually have quicker startup/recovery and have two hits, so they can blast through a regular one-hit fireball. In previous titles, you used a chunk of your super gauge to do this, but in SF6, it’s tied to the drive system. What’s new here are the meter management considerations rather than the actual utility of the moves (weighing EX move use vs. other potential uses of drive gauge).

Drive Reversal

Input: → HP+HK

Cost: 2

Compare with: Alpha counters (Street Fighter Alpha), V-Reversals (Street Fighter V)

While blocking, you can do an attack to push your opponent away to gain some space. They’re negative on block (universally -8 maybe?), but on hit, they knock your opponent down and reset to neutral. Because the input for Drive Impact is HP+HK, if you attempt a Drive Reversal and you aren’t actually blocking anything, you’ll get Drive Impact instead. You could potentially OS this deliberately, but this also might mean using your meter on something you didn’t intend.

Conclusion

It remains to be seen how all these systems tie together and what high level play will ultimately look like, but I think this is shaping up to be a very different game from previous Street Fighter titles. System mechanics seem much more at the forefront, so character-specific stuff might end up taking a backseat this time around. In general, it seems like a more complex game than we’re used to (especially when compared with something like launch SFV). With so many things being tied to the drive gauge, resource management is going to by key (I like this sort of thing enough to have made a whole video about it, but I'm not sure how it'll land with the playerbase at large). Obviously I'm also not sure how much will change between now and release—this will be an interesting(embarrassing?) post to revisit 6+ months after the game actually launches.

• High frequency producers: people who make content weekly or even multiple times per week. Think Sajam. They might make things like:

  • Reaction videos to recent FGC news

  • Tutorials teaching you the basics of how to use a character

  • Online matches with commentary (often streamed first)

• Low frequency producers: people who can have multi-month gaps between videos. Think Core-A gaming. High frequency producers usually focus on things like:

  • Documentaries

  • Video essays

High frequency producers make lots of videos, so they make videos quickly. This means they're in position to succeed where a fast turnaround is key. A character basics tutorial video illustrates this well. When a new character comes out, people aren’t going to wait months for someone to put out a meticulously researched, expertly edited, and well-produced tutorial video. Players just want to know how the character’s moves work, a couple good buttons, and a few basic combos to get started. There’s also a huge first mover advantage here. Whoever is first will get all the traffic (there’s no competition when you’re the only one out there). By default, they become the most popular video on the topic. Popular videos tend to be recommended by The Algorithm™, so the first mover tends to stay at the top of that list—it’s self-reinforcing. If you take your time and make the best character tutorial ever, you’ll likely never see the success of the person who made the quickest character tutorial ever. The most viewed Janemba tutorial video is one that came out the day the character dropped. Low frequency producers make few videos (and spend a long time making each one of them). Documentaries or video essays take a long time to make, so you naturally see low frequency producers focus on those.  

The Economics of Eyeballs

Whether making videos is your job or just a hobby, you want people to watch your videos (why else would you make them?). Let's say you make your living via YouTube and your goal is to get one million views per year. If you put out two videos per week, you’ll get there with ten thousand views per video. If you put out two videos per year, those videos need to average a half million views each. It sort of balances it out—if you’re spending more time making each individual video, each video should (hopefully) attract a larger audience. I don’t monetize my stuff, but I still think about this a lot—the time I spend making a video vs. the time people spend watching it and the value they get from it. If I spend a lot of time and effort making a video and lots of people watch it, it was time well spent. If not, it might be time to reevaluate.  

All of this is to say—if you’re interested in making videos, you need to think about what kinds of videos you want to make and what this necessarily implies. High frequency = high consistent time commitment. High frequency producers already exist for most (all?) major games. Some of these creators make videos full time. If being first means everything, how are you going to compete with someone who has already made it their full-time job? This becomes even more difficult when established creators get early access to new characters, game betas, etc.  Low frequency still takes lots of time, and while you avoid the constant time pressure of high frequency content creation, you need to offer something new and present it well enough to grab people’s attention. Neither path is easy!  

Happenstance Matters a Lot

Working hard will load the dice more heavily in your favor or let you roll the dice more times, but at some level, success will always rely on chance.  Examples: 

• Shoryuken.com used to exist and was an easy way to get videos in front of lots of people. Whether avenues like this exist for you is beyond your control.

• Your video might be mentioned or linked by somebody with a large social media following. Whether this happens is largely beyond your control.

• You might have started making videos at a time where there was more or less competition in the space. You can choose to enter a space where there’s less competition, but that probably means there’s less interest there in general. It could work out if the niche you pick ends up later exploding in popularity, but this is largely beyond your control.

Some amount of success in any creative endeavor depends on luck. You can take this as a positive or a negative. “I hate this, I should succeed purely based on my merits!” and “I failed, oh well, not my fault!” are both valid interpretations.  

On Production Value

Cool visuals don't matter as much as you think. This might seem hypocritical coming from a guy who spends hours making these:

I make them because it’s fun, but I think this sort of stuff basically has no effect on how well your videos do. That being said, some amount of production value does matter. Things like the intro sequences for Think, Don’t Mash! aren’t necessarily super time-consuming to create, but they’re clean and do their job (I hope, anyway). Editing, conciseness, and planning are what really matter. For example, take this Janemba tutorial video. It has a ton of useful information (I watched it and got a lot out of it!). However, it’s a 42 minute unedited stream of consciousness monstrosity of a video with no time stamps. You could put all the same information into a video that’s a third the length. I watched this particular video because I knew the player knew what they were talking about, but 99% of the time, if I click on a video and it has zero editing, I nope the hell out of there without a second thought. Speaking of unedited streams of consciousness, that's kind of what this blog post is turning into, so I think I'll call it here.

A while back, I wrote a blog post about what a simplified Guilty Gear game might look like. While anime/airdasher fighting games are fun, they are traditionally more niche within the competitive community, with fewer tournament entrants than series like Street Fighter or Tekken. I thought that this niche status was at least in part due to their complexity/difficulty (whether real or perceived), so I wondered if some kind of simplified “introductory” airdasher would help expand their audience.

For my thought experiment, I used Guilty Gear Xrd as a starting point, and then speculated about what mechanics you could remove to simplify the game (while still having the game feel like Guilty Gear). I wrote this before the release of Guilty Gear Strive (back when Strive was still being referred to as Guilty Gear 2020). At the time, we’d only just started learning about the mechanics of the game.

When I compared my Guilty Gear for Dummies thought experiment with what we knew at that point about Guilty Gear Strive, the two seemed closely aligned. Maybe that’s not too surprising—anybody looking to simplify Guilty Gear would likely independently come to many of the same conclusions.

Let’s fast forward to the present day. Strive has been out for almost exactly a year now. Is Strive the simplified version of Guilty Gear I’d imagined? 

Let’s not bury the lede. I think there are two points worth examining—

1. Guilty Gear Strive did simplify a lot, but I think they went about it in a pretty weird way, and they also made some things more complicated and less approachable. This is not a value judgment of whether it’s a good game or not, but Strive is not really simplified Guilty Gear.

2. I probably overrated the value of simplicity. Despite point #1, Strive has been a very successful game.

Strive—Not That Simple

Before release, it really looked like Arcsys was trying to make Strive more simple and accessible. We found out that they standardized wake-up timings, they allowed you to air-block ground moves without using Faultless Defense (FD), and they flat out removed more esoteric mechanics like Blitz Shield and Instant Kills. They got rid of Dead Angle Attacks and integrated their functionality into the Roman Cancel (RC) system instead, which instantly made sense and was a very elegant way to simplify without actually losing the mechanic.

However, some things went in the opposite direction. As soon as the beta tests began, everyone was frustrated with the gatling system. Charts were made attempting to lay it out for people:

You could argue that if you made the same kind of chart for an Xrd character, it’d look just as convoluted, but that’d be missing the point. In Xrd, like in most airdashers, weaker buttons can just chain/combo into stronger buttons (for the most part). 

For example, In Ultimate Marvel vs. Capcom 3, from weakest to strongest, your attack buttons are L (light attack), M (medium attack), H (heavy attack), and S (special/launcher). Most characters can just combo L > M > H > S. The timing is pretty lenient, and it’s easy to explain. Unlike Street Fighter, where you can get by without combos for a while, in airdashers, you really do need combos right off the bat, so making basic combos straightforward like this helps ease people into the game.

In Guilty Gear, the button strength order is punch > kick > slash > heavy slash > dust. Pre-Strive, you could start with your quicker/weaker buttons, and work up the stronger attacks to piece together a simple combo. This is easy to understand if you look at the default button layout on an arcade stick:

If you start with kick, you just drum your fingers across the top row of buttons from left to right and get a basic combo.

The Strive system is simpler in a sense—from a given button, you have just one or two options for what to do next, but you can’t summarize it in an easily remembered rule of thumb like, “just go to the next stronger button.” There are a bunch of individual special cases you need to memorize and internalize.

It seems like they wanted to limit the options in order to make offensive pressure less scary, but I think the method they used is inelegant and feels clunky.

Maybe Accessibility Doesn’t Matter?

If you made it here, you might be thinking, “Zissou—I get what you’re saying, but Strive seems to be doing really well.” You’re right! It’s not clear how well the game has sold to date and how that compares with their sales expectations, but the game is clearly very popular with players.

What’s going on here? Story time! A while back, I was talking with a friend who was into Smash (Smash 4 was pretty new at the time). I forget the exact circumstances, but it went something like this: 

1. I remarked that I wished I could play Smash at a basic level. 

2. He said he’d thought the same thing about Street Fighter. 

3. We agreed that for a couple weeks, we’d practice the other person’s game while we tried to coach each other.

I taught him Street Fighter V with him using Chun-li. I basically followed the plan from my Think, Don’t Mash! videos. On the Smash side, he recommended I try using Dr. Mario, since he was a pretty standard character so it’d be a good way to learn the game. He explained that short hop > neutral air was a strong tactic. If you’re not familiar with Smash, to short hop, you press and release the jump button really quickly (apparently you have three frames, i.e., 3/60ths of a second to do this). I found it pretty difficult to do consistently. 

Smash also has tilt attacks. These are strong grounded moves that you want to use pretty frequently.  These are performed by pushing the analog stick just a little bit in one direction and pressing the attack button. If you press the stick too far, you get a smash attack rather than a tilt, and I found it very difficult to get tilts when I wanted them.

People usually bring up Smash as an example of accessibility done right, so naturally, I wondered why nobody talked about how hard it was to do some of these very basic techniques. I think the answer is that nobody cares. In fighting games, the most visible and iconic moves a character has are their specials. When people think of Ryu, they think of hadoukens. When people think of Mario in Smash, they think of fireballs, and anybody can hop in and immediately easily chuck fireballs. Sure, they can’t do short hops or tilts, but it doesn’t matter to them. They don’t even know those things exist.

What actually matters is that people believe a game is accessible. The actual mechanical complexity of the game might just not matter that much. Arcsys convinced the FGC that Strive was a good entry point for the Guilty Gear series, and that’s kind of all that matters.

After the disappointing teaser trailer Capcom dropped earlier this year, I was surprised to find myself coming away extremely impressed after watching the recent Street Fighter 6 announcement trailer. What made me truly optimistic is something I’ve been thinking about a lot lately (which I touched on in my recent video): fighting game business models.

Fighting games have always had to contend with having to appeal to and satisfy two pretty separate audiences:

1. People focused on the competitive multiplayer experience

2. People who are interested in single-player content

Obviously there’s some overlap—there are plenty of people who enjoy both. However, I think there are huge portions of the fighting game audience who care mostly about just one or the other. This isn’t necessarily surprising, but it can put fighting game makers in a tough spot. You can choose to focus on just one of the two groups or you can try to please both, but each approach has its own challenges. 

Business Models: Three Paths

Let’s step outside fighting games for a second and look at shooters—a genre that has to navigate this same challenge. You see three main approaches in the shooter space:

1. Multiplayer only (only aim for audience #1):

• CS:GO

• Fortnite

• Apex Legends

2. Singleplayer only (only aim for audience #2):

• Half-life series

• Metro series

3. Both (which is really just two separate games inside of a single game to try and please both audiences)

• Gears of War series

• Call of Duty series

Multiplayer-only shooters are mostly free to play. This helps build a huge userbase (key for long-term player engagement) and with that large userbase, they can monetize in various ways (selling cosmetics or whatever). There are plenty of shooters solely focused on this space which do well. Fortnite banks billions in profit annually.

Singleplayer-focused shooters can charge standard retail prices and people are cool with that. Everybody is accustomed to spending $60+ on AAA singleplayer experiences. I would also put games like Doom (2016) in this category. While it has a multiplayer component, nobody really cares about it. The game’s success is solely due to its singleplayer campaign.

Games that do both, like the Call of Duty series, still charge standard retail price, and people are fine with it because the campaign is substantial enough that the $60 feels justified even if you ignore multiplayer completely.

There are a few games that fall outside this three category paradigm (like Overwatch), but I think it pretty cleanly describes most of the shooter market.

How do fighting games fit into all of this?

Back at the genesis of the genre, I think fighting game makers were mostly focused on the multiplayer audience, which makes sense for the arcade environment. Standards for single-player content were pretty low in the early console days (SNES/Sega Genesis era), that people were happy with ports of games like Street Fighter II, even though the singleplayer was usually limited to arcade mode.

For a time, fighting game makers did a pretty good job of satisfying both the singleplayer and multiplayer audiences with games like Soul Calibur, up into maybe the sixth generation of consoles (Dreamcast/PS2/GameCube/Xbox). But then, as we got into the PS3/Xbox 360 and PS4/Xbox One era, fighting game devs couldn’t decide which audience to aim for or what they wanted to be. 

I think Street Fighter V is the poster child for this identity crisis. It really seemed like there were people within Capcom with radically different ideas of what the game should be, and the game was a mess at launch. It really felt like SFV was supposed to be a free-to-play title, aimed squarely at the multiplayer audience, but then at the last minute, Capcom got cold feet and pushed it out the door as a full-priced $60 title, leading to a disastrous launch. It took years of work on Capcom’s part to shake that reputation and build SFV into a well-liked entry in the series.

I’d be remiss if I didn’t point to the one fighting game dev which has done a fantastic job with navigating the games market. With their revival of the Mortal Kombat series and the creation of the Injustice Series, NetherRealm Studios 100% committed to that third plan, building a full featured singleplayer experience worth $60 while also building a separate strong online multiplayer component. And they’ve been rewarded for their efforts—according to Wikipedia, Mortal Kombat 11 was the fifth best selling videogame of 2019 and has now sold over 12 million copies.

So why did I like this Street Fighter 6 trailer? There’s a lot to like, but personally, I think the single biggest takeaway is that Capcom picked a direction and committed to it. They’re clearly pouring resources into both the single and multiplayer pieces to truly justify it being a full-priced AAA release. I’m excited to see how it pans out.

My First Video

Back in June 2010, I was living in Japan, and the current home version of the game was Super Street Fighter IV. In the best of circumstances, I might be decent at fighting games, but back in 2010, I was pretty terrible at them.  I didn’t play SFIV too seriously, but I’d play with a friend here and there, and sometimes we’d take turns playing against random people online. One of these times, we happened to join a room with Sako, one of the most legendary Japanese fighting game players of all time. I lost the first match (pretty horribly), but through some kind of miracle (and because Blanka is a super scrubby character), I managed to win game two.  Sako immediately kicked us out of his room, and we sat there in disbelief. I needed to prove to people that this was a thing that happened. We set up a shitty digital camera on a GorillaPod and shot an offscreen video of the match replay, which was the first video I ever uploaded to YouTube:

I thought it’d be fun to have a jokey title, so (perhaps unwisely) I went with “Sako? More like sucko! America redeemed.” It turns out that sarcasm doesn’t really come across well on the internet and I was immediately bombarded with people defending the skills and honor of Japan with some very unkind words to say to me. People’s overwhelming hatred of Blanka probably didn’t help my case a whole lot. Oh well.

Marvel vs. Capcom 3 and Guilty Gear Xrd

A couple of years passed, and I’d become obsessed with Marvel vs. Capcom 3, and I only got more into the game when Ultimate Marvel vs. Capcom 3 dropped in late 2011. I liked SFIV, but I fell in love with Marvel 3. It was a game made for lab monsters, and I’d hunt around on YouTube and the Shoryuken.com (SRK) forums looking for new tech for my team (Doom/Dante/Strider). In mid-2013, I struck gold⁠—I found a video clip of a crazy Doom/Dante combo extension that looked like it provided a huge damage boost. The problem was that the footage of the combo extension was low quality, and it was impossible to tell how much damage the extension actually added to the end of the combo. After some testing, I realized the extension meant that you could basically kill any character in the game off of any starter. I threw together a quick video about the extension:

[embed]https://youtu.be/fJ3VE8GRja0[/embed]

This video was the first in a number of (very Doom-focused) Marvel 3 videos I put out on my Zissou YouTube channel. I was also very fortunate that at that time, SRK was a popular site for fighting game players, and they were happy to post my videos. This meant that instead of trying to scrape together a few hundred views, I could reliably get thousands. I also started playing Guilty Gear Xrd when it launched in arcades in 2014. I wanted to branch out to a new game since I was worried about the future of Marvel 3. I also figured I could get a head start—I could read Japanese well enough to get info from places like the Japanese wiki and top Japanese players’ Twitter accounts, and I had access to the arcade version of the game. This set me up to put out a bunch of Faust-focused Xrd tutorial videos over the next couple years.

The Marvel 3 and Xrd videos were pretty technical tutorial videos which were inherently going to appeal to a very narrow audience: Faust players in Xrd and Doctor Doom players in Marvel 3 (to be fair, there were a lot of Doom players). 

A New Direction

I eventually moved back to the U.S. not long after Street Fighter V came out, and I went all-in on making SFV videos for a hot minute. I tried some different types of videos, but I was really surprised/happy with how well one particular video was received: Street Fighter V - Defense and Dealing with Problematic Moves. Rather than a super technical/specific tutorial video, it was a broader video aiming to help players learn how to think and problem solve defensive situations in fighting games. It seemed obvious to me that I should keep going in that direction, and I thought it made sense to do that on a fresh channel— it was a pretty big departure from what I’d done in the past. Also, I’d used a lot of copywritten music on the Zissou channel, and I wanted to distance myself from that in case it ever caused any problems. I called the new channel Press Button Win.

Random Aside—The T-Shirt Business is Rough

I’d actually registered the pressbuttonwin.com domain name well before the new channel launch. I thought it’d be fun to design and sell t-shirts with fighting game-inspired designs. I put a short promo for the shirts on some videos on my Zissou channel. I was pretty regularly getting thousands of views per video at the time, so I figured if even 0.1% of my viewers grabbed a shirt, I could sell a few and help offset the costs of video capture equipment, editing software, and other stuff I needed for video production. Long story short: the t-shirt thing bombed completely. I did manage to sell a decent number of them in person when I had a booth at KVO one year, but online sales were basically nonexistent. I could talk more about what happened here, but I’ll save that for another post. Needless to say, I now have a lifetime supply of goofy t-shirts.

The t-shirt business (quickly) came and went, but I liked the name Press Button Win, so that became my new, more broadly focused YouTube channel.

The Path of Press Button Win

I really like the idea of teaching players how to learn, and I also wanted to convince people that with the right approach, anybody could learn and be at least decent at fighting games. I hunkered down and got most of the first four Think, Don’t Mash! videos finished before starting to release them so I could keep up a steady release schedule.  I got up through Part 7 of that series, and I felt like I’d pretty much said everything I wanted to say on the topic. I’d always wanted to make what I thought could be the definitive videos on a few intimidating fighting game concepts: frame data and hit boxes, so I went to work on those. Alongside those, I also made videos that leaned more in the video essay direction, like Reacting in Fighting Games and Customization in Fighting Games. The essays were the most fun to work on, and I think they resonated with more  people, which got me to the most recent video on the channel: Fighting Games—Punching Above Their Weight. I think this video essay approach is the direction I’ll continue to push (at least for the near future). 

I Can’t Believe You Read This

If you made it to the end of this post, congratulations! This is some self-indulgent nonsense, so I can’t imagine it to be that interesting for too many people, but it’s my blog and nobody can stop me. In my follow-up to this post (which will hopefully actually be interesting) I’m going to focus on the lessons I’ve learned about content creation along the way.

I've always thought fighting games could be more popular. I'm sure I'm not alone. If you're a fan of the genre, you've probably noticed fighting game devs seem to continuously make the same blunders and stubbornly resist change. I'm going to use this blog post to expand on some of the ideas in the video as well as throw out a few disclaimers. Watch the video first and then read on:

I'm fairly certain this will be the most controversial video I've made (though maybe this isn't saying much, haha). I hope the video itself does a decent job of speaking for itself, but I think it'll be helpful to add a bit more context and explanation here. Let's start with a couple quick disclaimers:

1. I don't hate MOBAs, first-person shooters, or whatever else. While I think they're harder to spectate and enjoy (especially as a non-player) than fighting games, they're great games. They're popular for good reason!

2. This overlaps with point #1, but while I used DOTA 2's "The Play" as an example of something really inscrutable, inscrutable doesn't mean it's not interesting, not hype, etc.

3. In keeping with the theme of the first two points, I wasn't trying to dunk on any specific fighting game. Games like Guilty Gear Strive and King of Fighters XV are useful as recent examples of some of things I talk about, but my points apply to nearly every fighting game.

This disclaimer gets to be its own paragraph. I'm guessing that the most controversial bit of this video is that I advocate for a free-to-play business model for future fighting games. First off—while I do believe this, I don't think every fighting game needs to do this or should do this. Also, I think there are a number of possible variants that are all worth exploring:

• The DOTA 2 model: DOTA 2 is completely free outside of cosmetic items. I think this is kind of the holy grail (where everything gameplay-related is free). However, this might be the riskiest model—if you don't get a big enough playerbase or just fail to sell enough cosmetic items, the game could flop hard.

• The League of Legends model: the game is free-to-play, with a subset of their characters (currently 16) being free at any given time. The free characters rotate weekly. You can permanently unlock characters with either real money or currency you earn playing the game. This seems less risky than the DOTA 2 model.

• The Overwatch model: Overwatch is not free-to-play, but it resembles DOTA 2: you buy the game once, but after that, all gameplay updates (new characters, maps, and modes) are free. All cosmetics can be unlocked through gameplay.

• The Halo Infinite model: The campaign is sold as a full priced product (launching for $60), but the online multiplayer component is free-to-play. I could see a Mortal Kombat title succeeding with something like this (given the strength of its single-player content), but NRS is so financially successful with their existing model that it's hard to imagine them making a shift like this.

It's not hard to imagine some hybrid of the LoL and Overwatch models. You could have a free-to-play version with rotating free characters, and then alongside that, you could sell a version for $60 that has all gameplay-related stuff included. This would allow players to stick with a more traditional fighting game business model while still making the game more easily accessible to new players with a free-to-play version. I'm honestly very surprised that Street Fighter V didn't do this. I don't think Capcom would've done it day 1 back in 2016, but I would've bet good money that some sort of free-to-play version of SFV would eventually be available. I'm convinced it was planned but then scrapped for one reason or another.

If you're curious, this was the original Reddit post that got me thinking on this topic, and this was my initial reply to that poster. I know that a month or two ago, the whole "Should fighting games be free to play?" question was the thing to talk about in the FGC for a couple days. Maybe my timing could've been better, but it takes forever to make the kinds of videos I make, so I hope people still find this video worth watching.

Anyway, I plan on updating this to address questions/comments as they come in on the video, so stay tuned!

I learned in high school English class that compare and contrast is the ultimate essay writing technique, and I can only assume nothing has changed since then. I think it's easier to understand the shortcomings of SF6's teaser by putting it next to a more successful reveal trailer. Let’s look at the Guilty Gear Strive trailer that dropped at EVO 2019 where people absolutely lost their minds. Take a second to watch this: Why did the Strive trailer succeed? First we have to look at context. Up until the trailer dropped, we weren’t even certain a new Guilty Gear was in development. People suspected it might be, but we had nothing concrete to go on. That’s the first big difference between the two reveals. Even if the Strive teaser was just a 10 second video with a guitar riff and the game’s logo, people would have been excited—Arc System Works confirming that the game was in development was new information. On the other hand, we’ve known for a fact that Street Fighter 6 is in development for a while now. Its existence isn’t news, and it certainly isn’t enough to carry a trailer.

Why did the Strive trailer succeed? First we have to look at context. Up until the trailer dropped, we weren’t even certain a new Guilty Gear was in development. People suspected it might be, but we had nothing concrete to go on. That’s the first big difference between the two reveals. Even if the Strive teaser was just a 10 second video with a guitar riff and the game’s logo, people would have been excited—Arc System Works confirming that the game was in development was new information. On the other hand, we’ve known for a fact that Street Fighter 6 is in development for a while now. Its existence isn’t news, and it certainly isn’t enough to carry a trailer.

The SF6 promo clocks in at 39 seconds compared with Strive’s 1:21. Comparing the two video lengths is kind of wild to me. While the Strive trailer is longer, we’re not talking orders of magnitude in difference here. In its 39 seconds, the SF6 teaser tells us almost nothing. It told us:

1. SFVI is in development

2. Luke is in the game

3. Ryu is in the game

As I mentioned earlier, we already knew #1. Capcom had also already stated their intent that Luke would be the face of SF6, so we knew #2 as well. While we didn’t know #3 with absolute certainty, putting five bucks on Ryu being in the next Street Fighter game is a pretty safe bet.

What did the Strive reveal trailer tell us?

1. Strive is in development

2. Sol is in the game

3. Ky is in the game

4. What the game actually looks like (showed in-engine footage)

5. The planned release year

6. A new mysterious bad-ass samurai guy is in the game holy shit

A few of these aren’t huge surprises (Sol/Ky being in the game), but everything else was new and exciting. I don’t think it’s really necessary to even dive into each list entry; it speaks for itself.

All of this being said, I’m not saying SF6 is going to be bad—that’s impossible to know at this point (we still haven't seen it). What I am saying is that this teaser doesn't seem to serve any real purpose. I guess the SFVI teaser did confirm hot Ryu is going to be his canon appearance in the game, so at least there’s that.

Hitboxes and hurtboxes! This is something of a follow-up to a previous video: Solving Problems with Frame Data. In fighting games, frame data is to time as hitboxes are to space. Together, they form the technical underpinnings that decide how fighting games interactions play out. Sometimes I use these blog posts as a way to introduce a video, but I’m going to use this one to expand on it instead, so it’s probably best to watch before continuing!

In the video, I stayed focused and talked exclusively about two kinds of (hit)boxes: hitboxes and hurtboxes. These dictate if and when hits actually land. If Balrog does a punch, it will momentarily create a hitbox that roughly corresponds with the location of his outstretched fist. If that box ever intersects the opponent’s hurtbox (which is usually roughly on top of their sprite/character model), the opponent gets hit. In this post, I wanted to at least briefly touch on the fact that there’s more out there than just hitboxes and hurtboxes.

Collision Boxes

If you look at a screenshot of Street Fighter V with the hitbox viewer enabled, you can see that there are more than hitboxes and hurtboxes. The green outlines show the characters’ hurtboxes. Inside those green outlines, there’s a blue outline filled with a semi-transparent yellow. This is called a collision box. The collision box governs whether one character can move through the other. For example, if you walk forward, once your character’s collision box touches the opponent’s, you’ll begin pushing them backward. When you see strange instances of characters passing through one another, it’s probably because one or both of the characters' collision boxes were either shrunken or absent.

Proximity Boxes

When you do a move in SFV, even before it can actually connect, it will force the opposing character into a block animation if they’re holding backward. This is called proximity guard. In the SFV hitbox viewer, the range at which a move will cause proximity guard shows as a yellow rectangle with no outline that extends along with the move. When this box intersects with the opponent’s hurtbox, they can no longer walk backward and the block animation begins. For this interaction specifically, the interaction triggers if the proximity guard box and the opponent hurtbox intersect on the x axis, even if the boxes don’t literally overlap.

If you imagine the hurtbox extended all the way to the top of the screen like a column, it’s easier to understand what I mean. These don’t have a widely accepted “correct” name, but since they’re boxes that trigger proximity guard, calling them proximity boxes seems pretty reasonable!

Conclusion

Some of these boxes won’t exist in every game. Many games don’t have proximity guard, so they have no need for proximity boxes. Some games might assign unique types of boxes to projectiles or armored moves. Whether you’re talking collision boxes, proximity boxes, or (insert thing here) boxes, the concept is still the same—they’re all just instances of geometric shapes governing how interactions play out behind the scenes. It’s fine to ignore all of them (and you can safely do so and be a great player!), but depending on your learning style, diving deeper into the world of hitboxes can help take your game to the next level. If nothing else, it’s fun to learn and understand what’s really going on under the hood.