Airdashing and Doublejumps

Air movement is a tricky subject. I got asked about this subject by our good friend like over a week ago, but I've not tried posting about it. I'm certain this will be a two or three post subject, surely.

The thing about airmovement is that it can really make or break things, depending on how things are implemented. A perfect example of both types of airmovement miserably failing would be Big Bang Beat. But possibly (depending who you ask) the greatest example would be Scarlet Weather Rhapsody, since it implemented almost-free 8-way movement in the air.

A good example of it actually working is probably Marvel vs Capcom 2 and Monster.

I believe the biggest issue with implementing either feature is how you implement various types of attacks and also guarding or airthrows in the air. If these are done poorly then it won't matter how you do air movement, the game will not be fun at all due to how linear and predictable things become.

The first thing that comes to mind is that SWR and BBB do not have airthrows, while standard and popular games do. Everything from Street Fighter to Melty Blood and Guilty Gear.

Now, SWR's predecessor game, Immaterial and Missing Power, also does not have air throws. But it also doesn't have 8-way air movement, only horizontal airmovement and no double jumps. I think that's a very important distinction.

Therefor, I'd like to first address doublejumps, and save airdashing for a later time.

The thing about doublejumps is they take the "rules" and a lot of the footsies mechanics that beat/break the "rules" that I was talking about in my previous post.. they take them and they throw them out the window. But only if the game doesn't support everything it needs to counter-balance doublejumps.

One golden-rule example is the jump-in. A player shouldn't just be allowed to jump at an opponent when both players are simply standing there doing nothing at around jump-distance. They should have to put the opponent in a situation where they are unable to anti-air by taking some measurement of risk. The other player should be able to anti-air them consistently and easily if they do not do anything strategical to achieve it. With double jumps this becomes a lot more complicated, you can jump again in the air right before getting anti-aired to avoid getting hit, and fish for anti-airs.

Another golden rule is risking a knockdown while moving forward. Dashing is obviously more risky than walking. But with doublejumps why bother with either when you can first move forward but jumping, then doublejump vertically and stay your ground, inching forward very safely.

These are major problems that designers have to deal with, balancing out things for both characters and game mechanics.

Fact is, if your character has a crouching light attack that can be rapidly spammed and works as an anti-air, while another character has only one valid anti-air that is a special move that takes them off the ground, things are unbalanced in this situation. Without doublejumps it probably isn't so extreme, but with them implemented then it definitely is polarized.

But, you can't just give the entire roster easy-mode light attacks that are anti-airs or that would defeat the purpose of doublejumps in the first place by destroying the point of being in the air at all. And if you're playing with balance by giving it to some but not others then you're inevitably fucking over the tournament scene unless you're a super genius (in which case you wouldn't be reading this) and can balance out every little thing.

However, it's not impossible to counter-balance doublejumps. As mentioned earlier, airthrows play a pretty big role. Another thing that really makes a difference is airblocking, and more specifically the lack thereof. But probably the biggest difference is the height of the doublejump itself and the overall priority/control air attacks have.

You could assume that 8-way movement is a sure fire way to fuck a game up, but it's actually hard to say. Jojo's Bizarre Adventure and MvC2 gave it to some characters, and SWR gave it to the whole roster. I'm of the finalized opinion that 8-way fucked SWR for good due to the terrible options for dealing with it. But it didn't exactly screw Jojo's and MvC2 entirely, and both games implement both pushblocking and guard canceling. It did, however, bring the characters that had this ability to the top of the tiers. Magneto and Kakyoin being the most celebrated in this regard. You could talk all day about how both characters have everything plus a bag of chips (Kakyoin, anyway), but I can guarantee you that Kakyoin isn't dominating just with his range alone.

Though, if you know anything about these games then you'd probably know that there's almost never any wiggling or ground-based footsies going on in them. Players will either rush in or fish aimlessly/wildly until they can begin the rushdown process, thus turning the game into a widely momentum based game.

I don't want to just go and say that it's an on/off switch though. As in, with doublejumps-footsies die. And the reason I don't want to say it is because Monster showed me that this is not the case.

The thing about Monster in particular was that doublejumps were very very short and low, they weren't actually a second jump at all but more of a little hop that changed your air trajectory. Additionally, even though Monster didn't have airthrows (cept for air cmd throws), it also did not have airblocking which alleviated most of the need for air throws (not entirely though, IMO).

You'd think that low doublejumps would be counterproductive, as in, I use to think that the extremely high doublejumps found in MB weren't problematic because anti-airs ended before the airborne character regained their position, but I was dead wrong and they sure as hell are problematic.

On the flipside, Monster's doublejumps weren't problematic at all except for how they affected combos. But for the most part, it didn't degrade or dilute footsies or how anti-airs worked. There was really only a few characters that had really poor anti-air options, the rest of the cast could deal with it just like any other game.

But this is where the on/off switch actually does occur, as far as I know. If a game is given 8-way air movement then it needs some ridiculously extreme defensive options to deal with it like pushblocking, guard reversals, instant blocking, assists, etc. If a game is given high double jumps then it turns into something like MB where players tend to hop around totally aimlessly, simply trying to stay above their opponent at the right time and fish for random hits.

In the end, the goal should be to preserve footsies while maintaining doublejump functionality. And frankly, most games do not, Monster being the only one that I can really think of that has both a strong footsie game and doublejumps implemented.

But even in games like GG you get matches where both players are hopping around in the air for almost no rhyme or reason other than to fish for random hits.

And that brings us back to the subject of airdashing, which I'll have to address later. Sorry for the currently scattered post, I'll have to revise it later as well.

Rules and Leading

For a lot of players there's a weird grey area in the basic fundamentals of fighting games. When we start out we tend to do things like jump and dash at the opponent a lot, but we quickly learn that human opponents can easily counter these things with consistency if their opponent does them blindly.

Some funny things happen after this, depending on the player(s). Some players get it into their head that these actions are invalid and stop doing them altogether. Some players are the opposite and keep doing them mindlessly. Some players start getting hit by jump-ins and get rushed down by midscreen dashes then wonder why they can't stop it, but the opponent can. And then some players realize the truth behind it, which is baiting.

These concepts (like anti-airs > jumping) are not set rules at all. A jump does not always necessarily mean that the person jumping should get hit. The reason is because players can bend and break the rules using other tools.


Leading is a word that I just randomly decided to use as a label for this particular subtype of baiting. It's generally just called baiting or footsies. But, I feel it's significant enough to pay it special attention, since it's very fundamental to almost every fighter.

I'll use SF3:3S for the sake of examples though, for whatever reason. In this case, Dudley and Ibuki. You don't really need to know much about these characters other than that Dudley has a very low and fast jump with good jumping attacks, but Ibuki has extremely good anti-airs and very low and long ground pokes. Thus, if Dudley jumps or dashes blindly then he will be hit on reaction very easily.

Therefor, leading is what Dudley must use to get in on an Ibuki player that is stationary and waiting to counter him. To do this, he has to walk within poke range of Ibuki and try to fish out or provoke a reaction out of the player. That means stepping into a suboptimal range and basically risking being hit at Ibuki's max range.

This is intentional though. By simply walking forward, the Dudley player has applied pressure to the Ibuki player, forcing their hand. Eventually Dudley can keep inching forward and Ibuki will have to take some sort of action to get him the hell away from her. Once that c.MK gets fired off, the rules change!

Ibuki's c.MK animates for 22 frames total (on whiff), and it's also one of the hardest things for Dudley to deal with on the ground against Ibuki, but it can also lead to Ibuki's undoing if properly fished for. Once it's out, Dudley is now "allowed" to jump. In fact, he should jump. Even if Dudley only noticed the c.MK and reacted physically half-way through the animation, Ibuki still won't be able to do anything until the c.MK is done animating, which won't be for another 11 frames (half of the full attack).

Now, 11 frames doesn't sound like a lot, but it most definitely is. Because Ibuki is locked in place there is no hope of walking backwards outside of Dudley's jump range, or walking forwards under him either. Ibuki also can't use her most reliable anti-air either, which is c.HP, because the startup time and positioning required for that is not possible at that point. This also limits exactly when Ibuki can even perform an anti-air, giving Dudley a much better chance of using an air-parry/attack OS. And that is how you change the rules.

This is the same with dashing, too. For the most part, most characters should never ever even dream of dashing towards Chun-Li when she has a full bar stocked and is just sitting there. But once you lead her into whiffing a laggy attack then the chance for her to connect a c.MK into super goes away.

This is why you see a lot of "wiggling" and "dancing" in games like SF3:3S. A lot of novice players just mimic it without understanding why the good players do it, but the above explanation is exactly why they do it. They want to pressure the opponent into slipping up and giving them an opening.

Throw Types and OS'ing

The subject came up regarding the different types of throws. Either a single button input plus direction, or a direction plus two-buttons.

There's quite a few significant differences in these types.

Whether by design or coincidence, single-button types tend to have very little or no start-up animation before the throw. Additionally, they almost never have a whiff animation either. Instead, with single-button type throws, if you attempt a throw when the opponent is not in a state or distance that can be thrown, you'll instead perform the attack associated with that button (usually a heavy punch or medium punch, for example).

Multi-button throws are usually very different, they come with some startup and also a whiff animation. What this means is that you'll very rarely or possibly never perform a throw by accident. If you pushed those buttons then you definitely meant to attempt a throw. And if you fail then you get a whiff throw animation.

Both have their pros and cons which effect gameplay rather dramatically.

First of all, in a game like Guilty Gear it opens up a world of pain. Guilty Gear throws are instantaneous, so you can even perform them as a reversal move and throw the opponent before taking damage.

But aside from that, there's also option selecting, which some players love and others loathe to death. What an option select does, is allow you to perform one action that results in the game picking the correct answer for you.

This would be like flipping a coin, and having it land on a mechanical hand that flips the coin to whatever side you called.

How it works with throws is, if you input an attack and then input a throw attempt immediately afterwards (before the attack occurs or even starts) the game will check to see if the opponent can be thrown. If so, the opponent will be thrown because the throw will cancel your attack attempt. If the opponent can not be thrown then the game will return a fail throw, which results in no throw attempt, and therefor you get your attack instead.

Thats how single button throws work anyway, since if you use just high-punch to throw, you'll do a highpunch if you can't throw or you'll throw if you can throw. But the option select allows you to change which attack you do, you could input a slash/medium-attack instead and also perform a throw attempt. That results in things like being able to anti-air the opponent if they attempt to jump, or throwing them if they stay on the ground, or hitting them out of a backdash if they attempt that, or even dodging a reversal/super move if that's applicable.

Multi-input throws come with another bag of pros and cons. With this type, depending on how lenient the input buffer for the game is, you may be able to perform an attack and then cancel it into the throw before the attack occurs/begins, like the single-button types. But in this case, it may result in your character being moved forwards, backwards, or even off the ground for however many frames prior to being canceled by the throw attempt. This occurs in games like SFA3, SF3:3S, and Monster. Possibly others.

Additionally, unlike the single-types, the multi-types allow you to perform the throw whenever you want, even if the opponent isn't currently in a state that can be thrown. Thus, even though there is startup to these throws, you can time the throw so that the active catch frames occur exactly when you want them to. For example, if you were to hit the opponent right next to you and then perform a single-type, you would have to push the button on the very exact frame that the opponent became vulnerable to throws. With multi-types, you can perform it in advance, which allows you to do things like meaty throws and throwing a jumping opponent right as they touch the ground.

But in the end, multi-types have a whiff animation, and therefor can be punished. Single types usually don't, or at least perform some sort of attack even when the throw fails, so they can only sometimes be punished, but can usually only be beaten in tick-throw trap situations and not punished at all due to option selecting.

Therefor, I was in favor of the multi-types for all of the above reasons.

Parries and All That Jazz

The final post in a series of posts regarding anticipation/reaction and beating/punishing.

These concepts are what has put the parry system in the fighting game Street Fighter 3: 3rd Strike under much scrutiny and ridicule, leading to much debate.

Fact of the matter is, a parry in 3S (and CvS2 for that matter) does not have a whiff animation. Therefor, you can't wait until the opponent is done attempting one to punish it. There is no time-frame between the act of a parry input and the transition to another action, which can include blocking.

You can, however, hit a parry during the actual attempt by hitting the parry during the input with a move that can counter a parry. What this means is that parries are beatable, but are not punishable.

What I enjoy using as an example of what this is like is a counter-move like Geese Howard, except if these counters had absolutely no whiff animation and did not put you in a state where you couldn't cancel the action.

In other words, if Geese performs a low-counter, it would have instant startup and instant recovery, and if at any time you connect a low-attack with the low-counter then Geese would be rewarded with significant damage. The only way to counter Geese's counter would be to hit with a high-attack to beat the low-counter.

Be that as it may, Geese is not in a state where he is unable to act, therefor if you throw him or perform a super attack that contains a super-flash and screen freeze, he has a chance to react and take evasive/defensive action.

And that is literally what a parry is. It is not like an uppercut/dragonpunch, and it is not a defensive option.

A parry isn't a defensive though. It's an offensive one.

A defensive option is like a Just Defend, or a Faultless/Fortress Defense, or a Push Block. Things that actually defend against damage without any direct followup/damage/knockdown. When something can be used to hit the opponent, and potentially knock them down, that's an offensive option.

A reversal dragon punch is an offensive option. Sure, you want to use it when the other person is attacking, but then you're attacking too. You're taking an offensive route to not just neutralize a threat but also to change momentum to your favor by choosing something that has a high probability of beating the opponent's attack.

A parry is the same thing. You aren't going to parry into nothing. If you succeed in a parry you're going to hit them and probably knock them down. This is an offensive option just like a DP.

Now, the difference is, a uppercut/dragon punch can be baited and punished on reaction.

A 3S-style parry can only be fished out and beat on anticipation.

And the primary reason is because it has no whiff animation. You can't hit a missed parry after it's executed, you can only hit an incorrect parry during the button input window.

If you're getting rushed down a defensive option will neutralize attacks, not reverse them. Offensive options reverse the flow of momentum.

But I digress. The end.

Beating and Punishing

This will be a follow-up to my last post.

Once again, it's so very simple, yet I continue to be shocked and amazed at how some players simply do not "get it".

What is a beat, and what is a punish.

For the most simple example I can think of, take a thief stealing a cookie:
- If you see a thief trying to take a cookie from a cookie jar, and smack their hand before they are able to remove it, this is a beat. You beat the race between the cookie to the thief's mouth and prevent it from being eaten.
- If you instead are unable to stop the thief from eating the cookie, but punish the thief after the cookie has been eaten, this is a punish.

Once again this effects fighters in dramatic ways. When the opponent is standing directly in front of you, there are some actions that they may perform that you simply can not punish even if you know they are coming. Meanwhile, there's some things that are beatable and some that are not, but the distinction gets a little blurred much like mixups and anticipation and reaction.

To put it in fighting game terms, let's take am uppercut or fireball. These moves generally leave the opponent wide open after they are performed. Uppercuts like Dragon Punches are often invulnerable on the way up, so there's really no way to prevent the opponent from getting that cookie, but you can punish them one the way down while they are recovering. On the flipside, fireballs are sometimes not punishable due to very speedy recovery, but you can sometimes beat them before the cookie even comes out if you are able to stick out an attack preemptively.

Reaction and Anticipation

This is a fairly simple concept, but it makes a huge impact on games. All forms of games.

But even with the simplicity of it, and the fact that it makes such a huge difference, it baffles me that there are many players who simply don't "get it".

For the most simple example I can think of, take a deck of cards faced down:
- Guessing what suit or color the card will be before the card is turned is anticipation. Therefor, even if it's an educated guess, it's still just a guess. You could not have possibly known what the card would be before it was turned, you called what it was on anticipation (preemptively).
- Identifying the card as soon as possible when the card is flipped is reaction. This isn't necessarily a guess, but if you press yourself very hard you may end up so scatter-brained that you guess anyway. However, you did at least get some time to check and confirm what card it was on reaction (actively).

How this effects games should be obvious for traditional card games, dice games, and other such games. How it effects fighters is actually split many ways.

First of all, what we refer to as mixups is simply the act of trying to hit the opponent with one of two to four options, each option having a right or wrong answer.

For the sake of argument, let's say there is always three options both players can choose from. Now if we remove reaction and just rely on anticipation, we can then say this is exactly like the classic children's game "Rock-Paper-Scissors". This is because you can not guess which option to choose after you have seen what the opponent has chosen, you can only choose in advance, which means a guess.

If we add reaction into it, the game becomes quite a bit different.

This is where some people get a little confused, I'm sure. They believe the guessing game to be a mind game, such as psyching the other player out with some sort of hidden Jedi-mind-tricks or somesuch. However, what is more commonly become known as a mind game in this sense is nothing more than a simple guessing game just like "RPS". A real mind game is not so simple, it has more to do with risk/reward sacrifices and reading (also known as Yomi).

Regardless, let's look at the second step. The second step is that Fighting games implement both anticipation and reaction on a blurred set of rules and with variable exceptions. When an opponent performs a mixup, it may sometimes be something slow enough to react to, but other times it may be something so fast that no human could possibly react to it and therefor it becomes anticipation-only.

It's interesting how some games go to great lengths to tone down the ability to perform mixups that force anticipation-based situations, and how players go to great lengths to study/modify/master/control/perfect and explore any and every form of mixup they can get their hands on that forces such anticipation-based situations.

RPG Items

The subject came up of collecting and conserving items in RPGs. Expendable items are problematic because it's in a lot of players nature to not use such items in order to save them for when they are really needed, but then these players end up never using them in the first place.

My first thought was to eliminate items in the traditional sense. Rather, items would be crafted/created by spending mana/spell points or possibly combining material.

Ideally the spell system would be re-worked to compensate for such a mechanic. Such as, some enemies would reward with more mana after eliminating them than others.

Then, in theory, the player could choose to cast magic during a battle to fight and then after a battle in order to heal, which would drain mana against tough opponents that don't reward with much mana. Or, the player could choose to conserve mana by fighting weaker enemies that reward lots of mana, then use the excess mana supplies to craft magical items to be used in battle against tougher enemies.

Some mechanic like this would also take away some of the random-drop factor with traditional items, and would put a curve on the use of currency for expendable items. Grinding in certain areas in order to fight certain monsters that drop specific items random is not very fun, it's a grind. But with a crafting system the player could be given a certain level of control over what items are made. Very useful items could be balanced with some form of value system. That would also enable players to choose to craft a very mana-expensive item, banking on the risk that they won't be needing mana for a while until they regenerate from the cost of creating a heavy-duty item.

In order to further encourage item use, items could be given bonuses to regular magic. Such as, magic could be very standard, but magical items could have the property of gaining potential critical hits or having some kind of affinity bonus.