How do I learn to play fighting games proactively?
Hi Patrick,
Hope you’re doing well. The title says it all, really. I’ve put time into several fighting games now (currently on SFV), and I’ve hit a similar plateau of sorts in all of them: If opponents don’t consistently give me their risks for me to react/counter to, I totally fall apart. I’m almost never the player “causing the problem,” unless I’m panicking, where I usually cause predictable problems.
There’s likely some emotional stuff happening mid-match that I’m not totally aware of yet, but for now, I’d like to hear your thoughts on this if any. Thanks a bunch in advance.
P.S. I love your “Relax, you’re doing fine” essay. Really helps in a pinch!
Sincerely,
Other Than Attack
Hi OTA,
Thanks for writing in! This is an excellent question that can be tackled in many ways, so I’ll write this with an eye towards all the different ways I’ve seen people work through this before, and we’ll see if any of this lands for you.
People usually come to fighting games with a predisposition towards either proactive or reactive play; in other words, whether you like to look at the screen and then do something, or do something and then look at the screen. (For a more detailed explanation, check out this real good essay I wrote on OODA loops. Seriously, it’s like one of the best things I’ve written about fighting games and no one gives it any love.) You’ll have to learn how to do both.
Fortunately, the plateau you’re describing is fairly common, especially in games like Street Fighter where the first lesson you learn is “don’t jump, work on your anti-airs”. You can get a decent amount of mileage just doing things like consistently anti-airing, safe jumping DPs and punishing, and punishing unsafe moves on block. That’s all good, but none of that is really playing much of a two-player game. Eventually you’re going to run into players who don’t do stuff that is easily beaten on reaction, so you have to learn how to force people to do stuff that you can beat.
The specifics of how and when and why you do this depends on the game and character, so instead of just listening to me, I’d recommend looking up some match videos to see how good players force action. However, I do think there are some general truths about learning to attack that are worth covering, and some of it starts by covering the realities of the genre itself.
You cannot win off of reactable play alone
I get a lot of low/mid-level fighting game players who proudly tell me that they like to zone people out and win from range, and I totally get this. Keeping someone out feels good. Forcing people to run into your pokes and projectiles and anti-airs while you calmly dispense pain from your preferred position is a great way to feel smart and cool while playing a fighting game. There isn’t a fighting manga out there that doesn’t have some kind of impossibly cool genius who spends three episodes just grinding down the hotheaded protagonist into a dead stop. And if you’re still getting used to the speed of fighting games, playing at midscreen and further ranges is generally more comfortable because everything at that range is easier to react to than up-close situations, where you’re often forced to make strike/throw guesses and you can easily get overwhelmed by the pace. I started my competitive fighting game career playing Capcom vs. SNK 2 and Marvel vs. Capcom 2 in Northern California, a region known for calm and measured neutral play, and that kind of playstyle felt incredibly satisfying to get good at, because it feels like the opponent is just slower than you.
However, that kind of playstyle also meant that Capcom vs. SNK 2 tournament finals took forever, and Marvel vs. Capcom 2 matches involved listening to two Storms screaming “OH OH OH OH” as they flew to the top of the screen while whiffing HPs to safely build meter for 90 seconds. Make no mistake, zoning play requires skill and can be very exciting to watch, but it often doesn’t leave the best first impression on people who are interested in seeing if they’d like the game, or on new players who don’t yet understand how to beat zoning and are at risk of simply giving up and calling the game shit instead of learning how to deal with it. So, just like how the SF4-era of fighting games gave us comeback mechanics all over the place, the SFV-era brought us all kinds of different tools and mechanics to bias the game towards the attacker. Guilty Gear’s Negative Penalty is the most classic example (and far predates SFV, of course); sure, you can zone in GG, but if you aren’t moving forwards, you aren’t getting meter, and if you don’t have meter, you’re unable to do most of your character’s coolest, most powerful stuff. SFV did this with mechanics like Crush Counter, which creates significantly skewed payoff opportunities for callouts with big buttons, but it also achieved this effect through its attack speed timing, narrower range differences in normal attacks, and built-in input lag that made whiff punishing go from a basic aspect of neutral play to a fairly high-level skill test. Instead of focusing as heavily on reaction challenges as they used to, modern fighting games keep the big payoffs behind some weighted guessing games (usually referred to as the rock-paper-scissors part of fighting games), so you have to make sure you’re playing that part of the game, too.
I think there are two things that we can do to get you a bit more comfortable with your scrappier side. First part is going to involve breaking down some of these games into basic weighted risk/reward skill tests, and the second part is going to be talking about creating advantages. Let’s start with the first part.
Fighting games: Because “Varied real-time weighted risk/reward skill tests” doesn’t have quite the same ring to it
In a fighting game, you win by reducing your enemy’s health to 0 before they can do the same to you. In order to reduce the enemy’s health, you have to hit them with attacks that deal damage when the attack’s hitbox intersects with an enemy hurtbox. However, if the enemy is blocking your attack successfully, they will not deal damage. Blocking is instant, and pretty much everything else has startup and recovery time. So, if you’re going to hit your opponent, you have to get them while they’re moving or attacking, trick them into missing their block, or do something that beats blocking (like a throw). Broadly speaking, “hitting them while moving or attacking” is basically “playing neutral”, “tricking them into missing their block” is “mixing someone up”, and “beating block” is “pressuring with strike/throw”.
Punishing people for doing reactable high-commitment stuff in neutral, like jumping in too much or throwing poorly spaced fireballs or whatever, is an important part of playing a fighting game, but in most games these days the reward for winning neutral by reacting to your opponent’s moves is not big enough or consistent enough that you’ll be able to win against a decent player just by doing this. The payoff for winning neutral isn’t Big Damage — it’s the knockdown advantage that you can use to apply strike/throw pressure or go for a mixup to get the Big Damage. (Having played older fighting games where just whiff punishing pokes in neutral on reaction was a viable strategy: there are few things that feel quite as futile as playing against someone who feels like they can hit you every time you walk forward on reaction. When games are built to favor reaction speed too heavily, you run into the limits of your specific human biology real quick.)
The thing about reactable situations is that they’re the “easy” skill check in a fighting game; all you have to do is see the situation, recognize it, and apply the correct response. If you put someone in a strike/throw pressure situation, then you’re forcing someone to guess what you’re going to do, in a situation where the risk/reward is in your favor. And if you put someone in a mixup situation — say, an ambiguous left/right crossup safejump — then you’re applying a knowledge check (do they know that this is a safejump and don’t try to wakeup DP it) as well as a guess (left/right crossup, maybe even high/low if you can go for an empty jump low in the mixup as well). If you’re looking to win consistently in fighting games, forcing the opponent into situations where they have to pass certain knowledge checks just for the privilege of having to guessing right is a great way to do it.
Of course, this doesn’t mean that winning in these reactable situations isn’t useful! But if your mixup and pressure game is good, then you only need to out-react your opponent one or two times a round instead of ten times, which is important because good players won’t give you something reactable to beat ten times every round. One of the funniest things I’ve seen new players do is actually do something good, like a clean anti-air or a whiff punish, and then back off to reset to a comfortable range so they can do it again. They’re feeling good about doing the good thing, which is great! But unless that good thing paid out in at least 30% of their opponent’s health bar, they’re leaving the real payout on the table by backing off instead of taking the knockdown advantage. It’s understandable, of course — maybe they don’t know a good mixup or they don’t have a combo to threaten with — but it’s clear that they don’t yet have a full picture of the game in their minds.
This can feel really uncomfortable at first, because you’re going from a gameplan where you don’t really risk getting hit to a gameplan where you’re putting yourself on the line. But with most fighting games, you’re going to win a lot more once you’re willing to bet a little of your health to take a lot of your opponents’ health, and that’s exactly the kind of situation you force your opponent to play in every time you knock them down and give them some mix to hold. So if you refine your knockdown play, you’ll get yourself to the point where punishing the reactable stuff is the beginning of your game rather than the end.
Now let’s get to the second part.
Creating frame advantage out of nothing
When you say “I’m almost never the player causing the problem”, it reminds me of how it feels to play against a player who simply outclasses me in neutral; I always feel like I’m operating a disadvantage, just barely escaping from things they do over and over until I’m making foolish decisions to attack or I’m backed into the corner. This is because they’ve literally put me at a frame disadvantage in neutral.
We’re used to describing frame advantage in hit/block situations, but it’s a useful concept for describing neutral, too. Your goal in neutral might be “to hit the opponent”, but when you pick apart how you get to do that reliably, a lot of it works by tricking your opponent into thinking you’re going to do something, baiting them into doing something bigger, and then punishing them for doing the bigger thing.
For example, let’s bring it back to some good old Super SF2 Turbo. Let’s imagine we’re playing the Ryu mirror and we’re standing just outside sweep range.
Ryu’s Crouching HK sweep has 4f startup, 6f active, and 25f recovery; HP Hadouken has 11f startup and 43f of recovery; LP Shoryuken has 4f startup, 14f active, and 26f recovery; Jump forward takes 50f total with 7f of cancelable landing recovery. Human reaction speed varies based on person and context, but we can generally ballpark it at around 16f-18f to recognize a cue that you’re specifically looking for and begin to execute an input.
At this range, fireball beats sweep, early jump beats fireball, and late jump loses to fireball -> DP. In this situation, the first goal is to get the opponent to jump at you because they think you’re going to throw a fireball, since early jump beats fireball. However, you’re close enough that they’re not jumping the fireball early on reaction; it’ll take them ~18f to see the fireball is coming and input jump, plus another 4 to leave the ground, at which point the fireball has been traveling at blistering speeds for 15f and probably hit them before they even left the ground. So you have to do something to get them to jump first, then DP them on reaction; something convincing enough to look like you’re throwing a fireball to get them to jump, but something low-commitment enough that you won’t get whiff punished. Usually people pick whiffing something like a standing LK (5f startup, 8f active, 5f recovery) which isn’t going to get whiff punished and kinda looks like the fireball startup.
If it worked, they spent a jumping HK (probably something like 30–40f depending on how deep the attack is) on your 18f feint, which is plenty of time to react with an anti-air DP. That’s a frame advantage of at least 22f! It’s not the same as hit/block frame advantage, but that’s still about a third of a second where they’re stuck following through on a decision they made, and that’s time you get to recognize it and punish them. If it didn’t work, you might have given them 18f to recognize that you did a feint — not free, but due to the time it takes to recognize the feint and react, they probably aren’t punishing you on reaction. If they instead bet that you were going to feint, however, and decided to throw a fireball at the exact same time, you’ll be lucky if you recover in time to block it, but even blocking will give them a significant frame advantage they can use to close the gap and threaten you further if they choose.
See what we’re trying to do here? Yeah, we’ll take a situation that we can react to the opponent’s jump and punish them if we can get it, but even if we don’t get something that big, just putting them in situations where they are committing to bigger decisions than you are, whether it’s walking forward after you see them whiff a poke, will give you a little bit more time than they get to make a decision. Sometimes you can use that frame advantage for actual damage, but sometimes it’s just the thing that lets you get a poke out while they’re recovering, or lets you dash up a little bit closer to force a strike/throw guess instead of poking, or whatever. The more you can get them to react to the stuff you do with bigger stuff than what you’re doing, the more time you get to play the game while they’re waiting for their turn.
If this sounds hard and abstract, well, it is — but that doesn’t stop you from learning how to do it, just like it hasn’t stopped anyone else. It’s good that you have the anti-air DP, now you need to think about how to throw fireballs in a way that makes them want to jump, then think about what you’re going to do after they get DPed and they’re scared off of jumping, and so on, and so on. Or if someone has a range they like to fight from, it’s on you to figure out how to exploit that preference by predicting that they’ll want to get to that range and choosing to do things that stop them from getting there safely. If you can get a read in neutral on what threat your opponent is looking for and what they want to do about it, you can create frame advantage without even having to touch them. And if your opponent is doing that to you, then it’s on you to identify where they’ve called you out and shake them off your thought process so you can get back to looking for an advantage. You can simply start out every match by asking yourself what you can do to get your opponent to jump or dash at you, decide on an anti-air and a poke to keep at the front of your mind for the next two or three seconds (that’s 120–180f!), and then go from there.
This is a workout for your brain
I know this is a long one, but one more thing before I close this one out: Do not worry about your wins and losses while working on anything, but especially this stuff.
What you’re doing here is the complete set of work involved in learning to play a fighting game. Learning the mixups and strike-throw pressure strings and turning them into cycled patterns is both studying the game’s combat systems in depth, and learning often very specific timings in order to achieve the optimal advantaged outcome. Practicing fighting for frame advantage in neutral is essentially breaking down the cognitive processing involved in a fighting game — performing a constant cycle of taking in new visual information, turning it into a game state, deciding how to play that state, and executing the decision, all while factoring in longer-term decision-making patterns for both yourself and your opponent — and learning to be more nimble and deliberate with how you do each step of the loop. It’s going to be exhausting and you will always be tempted to switch to the cognitive patterns that you’re the most comfortable with.
When it comes to this kind of practice, just doing the thing as much as possible is your success condition. It will likely be hard. But trust me — when you get better at this stuff, it feels like the game itself slows down.
For what it’s worth: I can relate pretty heavily to your plight. Not only is this similar to where I first found my footing in fighting games, but it’s also my comfort zone for martial arts, too. When I’m doing actual boxing/kickboxing sparring, I tend to prefer staying on the outside and going for easy low-risk jabs and low-commitment kicks to create openings because a) I’m tall and lanky enough to do it and b) I’d rather get one hit in and take zero damage than get three hits in and risk taking some damage. (Gotta keep the moneymaker looking good, you know?). But when I spar with people longer than I am, I can’t rely on my length advantage any more, and when I spar with people who are just better at striking than I am, I know that my narrow comfort zone makes it easy for them to identify what I like to do and beat it. It’s not wrong to have a way you like to fight, but if you’re not comfortable and competent playing every aspect of the game, then it makes it that much easier for the opponent to figure out what you don’t like to do and keep you there.
Thanks again for writing in!
—
Thanks for reading! If you found this essay valuable and want to support my work, please do not hesitate to share it around on your social channels, follow me on Twitter, check out my Twitch stream, or join my Patreon.
💪😎👍❤
-patrick miller
