Combos: Length, Versatility, Interaction

Ever since their unintentional inclusion in Street Fighter II, combos have been an integral part of fighting games, and one of the most telling ways of how design philosophies in fighting games differ. Nearly every single fighting game has a different combo system, and learning to harness and exploit it can be one of the most satisfying things to do as you learn a game. On the other hand, games in which long combos are standard are often decried for a lack of interactiveness on the part of the victim, as, by definition, the victim cannot perform any action while being comboed, barring game-specific mechanics.

Throughout this blog post, I will discuss the combo systems of a few different games, and in particular, how these combo systems interact with the rest of the game. Combos, and their relative importance to how a game plays, both reflect and shape a lot more of the game than is immediately obvious.

In general, combos always serve the same purpose: to remove a chunk of the opponent's health, and to move them to a position where they are still disadvantaged after the combo ends. Depending on the game, combos may also play an important role in building meter for use in supers, EX moves, or other game-specific mechanics. The interesting part of how combos differ between games lies in how different combo systems achieve this goal.

Two opposing systems: Street Fighter and Blazblue

We'll start by looking at Street Fighter IV's combo system. It's very heavily based on links, sequences of attacks that are performed in full where the opponent cannot act in between attacks. This is differentiated from cancels, where the ending portions of a move are canceled by starting another move; Street Fighter places relatively low emphasis on cancels. There is an element of mechanical difficulty here, since the game does not include an input buffer system. Often, standard combos involve timing attacks to a single frame. A discussion on the value of mechanical difficulty is beyond this post, however.

An old combo tutorial for Ryu. The combos have changed but the principles remain the same.

Watching the video above, you can see a few common patterns. First, combos have very few hits outside of heavy resource usage. This is a direct result of the link-heavy nature of the game and the relative immobility of the character. Second, the standard combo ender is a sweep that provides hard knockdown, at near-max range. This is an important note in Street Fighter: you almost always want to end a combo with a hard knockdown. Third, each hit tends to come very quickly, with very low startup and low hitstop when the attack connects. We'll explore that later in this post.

I want to expand on the second point for now. In this game, combos tend to end with the attacker on the ground, knocking down the victim. Since characters have relatively few mobility options, this leads to a low variety of things to do as okizeme, meaning attacks performed as the opponent is getting up to try and put them in another combo. You can retreat to play more of a spacing game, you can time an attack to hit on the first frame the opponent is vulnerable (a meaty attack), you can throw, or you can try to do a high-low mixup with an overhead (often not leading into a combo) or a left-right mixup with a forward jumpin attack.

Relatedly, the major point of any combo is to either knock down the opponent or to just KO them outright. There's very little else to think about when performing a combo, except whether or not you want to spend your meter resource to do more damage or get a knockdown you would otherwise be unable to get.

You can see the result of this philosophy, along with what happens when you make a character that exploits that philosophy, in this match from this year's Evolution tournament.

Infiltration vs. Daigo is a great exhibition of what happens when a character can exploit the combo system.

Throughout the match, you can see that both players' combos end in knockdowns almost exclusively. Akuma, especially, prefers to get a knockdown and then jump at the opponent, doing one of myriad mixups made available to him because of his character's strong air options, ending that combo in another knockdown and repeating the process. This behavior is known as a vortex of mixups, and a large number of characters new to Street Fighter IV have strong vortex capabilities, such as C. Viper and El Fuerte. These characters are often disliked because strong mixups have not traditionally been part of the series' philosophy--instead preferring strong spacing and getting single hits out of this spacing to convert into high damage.

The low versatility in combo ending is what makes vortex characters like Akuma so powerful. You only really have two ways to end a combo: in a knockdown at the edge of attacking range, or with the opponent standing (for suboptimal combos or situations). So any character with a strong oki game will be able to exploit this low versatility and maintain their offense until their mixup is successfully blocked. If there were more ways in which combos could end, then selecting different vortex options could restrict further mixups.

Vortex characters aside, this system importantly makes it obvious what the focus of the game is. What combos do before their knockdown does not typically affect the attacker-defender relationship; as long as they end in the same way, what happened before that doesn't matter. In Street Fighter, a combo is just a mechanic to maintain the attacker's advantage upon landing a hit. Everything else is damage filler.

I'll go into my opinion on that later; right now we need something to compare that combo system to. So next we'll examine Blazblue. This game's combo philosophy is hard to pin down--characters are very asymmetric compared to Street Fighter and have vastly different methods of performing combos. The only thing that remains mostly constant across all characters is that combos outside of the corner are generally short and low-damage, and combos in the corner are very long and very high-damage.

This Litchi combo tutorial exhibits a lot of the decision-making necessary when comboing in Blazblue.

You don't need to watch the whole video to get the important points. The most obvious being that there are a huge variety of combos from a huge variety of starters. This itself would be meaningless if it weren't for the fact that the high versatility in combo selection allows a player to choose from a variety of variables to focus on in any given combo. You can prioritize damage, meter gain, corner carry, resource usage, type of finisher (and therefore the oki you can perform off of it), and/or difficulty.

Your sheer number of choices in combo has an interesting side effect in that it colors your neutral game, as well. The starting hit of the combo greatly limits your choice in how you perform the remainder of the combo (due to the underlying mechanics of the combo system, which will be the subject of a later post), so you will want to choose which of the aforementioned variables to focus on and try to force attacks to connect that enable combos that do well in those variables.

Looking at this from the defensive player's point of view makes the yomi component immediately obvious--with knowledge of the opponent's combo options, you can try to discern what the opponent wants to prioritize, and pay special attention to defending attacks that can lead into the combo the opponent wants. While being comboed, you can see how the combo is progressing and try to determine the kind of oki the opponent is going for, making your defense better once the combo ends.

On the other hand, the long combos mean that getting hit is a very non-interactive experience. There is little that the victim can do while being comboed; your only option is to burst, which will blow the attacker away, ending the combo, but bursts can be baited, blocked, and punished so there's a yomi component there, if a small one.

There are two huge losses in this system. The first is that with the game being so combo-centric, the traditional fighting game focus on spacing and footsies doesn't apply. To facilitate versatile combos, characters have extremely powerful movement options, particularly in the air; characters can double jump and airdash, as well as block while airborne. Such strong movement renders obsolete most of the ground-based spacing game that many fighting game fans love. Additionally, attacks are generally slower and have more hitstop, making hit confirmation and combo improvisation a lot easier, at the cost of perceived pace. Blazblue is a game of movement, mixups, and combos, not of spacing, which is generally not liked.

The second huge loss is in accessibility. Full exploitation of the combo system requires intimate knowledge not only of your character, but of the opponent's. You can't assume your opponent's combos will all end in a medium-range knockdown, and your opponent's oki options will vary wildly based on the character they play. On the offensive side, playing without a strong repertoire of combos is a massive disadvantage, much more so than in typical ground-based games. The knowledge floor is very high.

Comparing the merits

While there are certainly other combo systems in popular games, right now I'd like to compare the benefits and drawbacks of these two games' systems particularly. In both games, standard combos do almost the same amount of damage relative to total health. Both games prefer to end combos in strong oki situations. In this regard, the end result of combos in each game is the same--but how they get there, and the effect these differences have on the gameplay, is worth discussing.

The most striking difference is, as mentioned, how combos shape the neutral game. In Street Fighter, combos are just the natural result of getting a hit; the shape a combo takes is not particularly important. It will do some damage, and end in a knockdown from which the player can do the same few kinds of mixups. This leads to a very ground-based spacing game in which combos are unimportant compared to getting the first hit. Indeed, the entire game is based on getting that single hit--the combo is mostly there to provide you with your knockdown afterward. This system is heavily favored by people with a more purist outlook on what a fighting game should be: a battle of spacing, yomi, and well-defined, constant options that persist across many situations.

Contrast this to Blazblue and see how it flips the relationship on its head. Combos are not the result of getting a hit; the whole point of getting a hit is to get a combo. The distinction in priority is subtle, but important in how it completely changes the design philosophy. Combos are the integral part of the game; there is a huge amount of tradeoff and decision making necessary when doing all but the simplest combos, and both the attacker and victim must be aware of it in order to get a proper read on each other's intention. Much of the game's yomi comes from figuring out what your opponent wants to do with their combo: how they want to start it, what they want to get out of it, and what they want to do after it. The entire neutral game is built around facilitating this versatility, with strong mobility and slower, more meaningful individual hits that put the opponent in a position to be hit by a variety of other attacks. This destroys a lot of the Street Fighter mindset by all but removing ground spacing, and allowing combos to end in myriad ways of varying advantage, with few constant, reoccurring options. Clearly, this system is favored by those who care less about fighting games as a pure expression of yomi and more about having strong, versatile options available to them.

For each system, one thing that the system gains is implicitly something that the other system loses. Street Fighter loses strong mobility in order to emphasize spacing; Blazblue loses spacing in order to emphasize mobility, etc. From a purely theoretical standpoint, there's nothing inherently better about either of these tradeoffs, except for which tradeoff brings you, the player, closer to your ideal for what you enjoy in a fighting game. Most players prefer the pure expression of yomi.

But Street Fighter, specifically, loses something that it doesn't need to lose. I mentioned earlier how Blazblue has a high knowledge floor because of how versatile combo options are, and importantly, why the versatility of combo options matters. Street Fighter IV, for some bizarre reason, has a massive number of character-specific combos, which simply does not need to happen in a game where the middle part of a combo is just damage filler. For a time, the game even had at least one costume-specific combo, though this of course was a bug and was fixed. Character hurtboxes don't need to be so vastly different that any given combo only works on some seemingly-random subset of the cast. Everything else I've mentioned thus far is, I believe, a viable tradeoff to implement differing, valid, design philosophies. But Street Fighter IV's character specific combos are, as far as I can see, completely worthless and unnecessary, and add a knowledge burden that isn't offset by any meaningful gain.

You could also argue that the lack of a buffer system, and the resulting one-frame links, are an unnecessary mechanical burden, but that's a subject for a later post on mechanical difficulty in general. Personally, I think mechanical difficulty can play a useful role outside of the obvious one of making certain options impossible or take a minimum amount of time to execute (see: 360s), which I'll expand on in that post.

Other games: Marvel and King of Fighters

Hoo boy, it's time to talk about Ultimate Marvel vs. Capcom 3, without a doubt the single most combo-centric game that Americans actually play (see me in two years when Under Night In-Birth makes it over here somehow). Marvel's combos actually make me angry. The game takes the two systems I compared earlier and combines them in a way strictly worse than either of the two alone. Like Blazblue, characters have extremely strong movement and attack options that lets them do very long combos that could end in a number of ways, but like Street Fighter, you're going to end almost all of your combos in the same way regardless, except rather than knockdown, it's character death. The end result is a set of mechanically-difficult motions that are just a binary pass/fail for whether or not your opponent's character dies.

Watch what happens when Zero hits you. He either makes a misinput, or you die!

Zero and Vergil are perhaps the two most reviled characters for their combos. Both of them do extremely damaging, meter-efficient loops that kill their target unless they're dropped. And as long as you have the meter to kill your target, the right answer is almost always to spend it. There is so little decision-making going on with Marvel combos that their length and power is indefensible. Every hit of a combo should be meaningful. In Blazblue, you select a combo based on how much you favor any of the variables mentioned earlier, and each hit in the combo exists to facilitate that. In Street Fighter, combos are relatively short because they're just a vehicle to a knockdown, unless extended with resource usage for the purpose of additional damage. Marvel has no such decision-making or justification behind its long combos. Combos are powerful enough to kill the opponent off of one touch, and you virtually always want to do that.

And once you've killed the target, you end up in the Street Fighter-esque situation where your opponent's next character comes in, at exactly the same angle from exactly the same position, and you have your super-strong mixup setups to get you your next hit and kill. Even combo ending diversity is destroyed in Marvel. In a recent blog post, former Evo champion Viscant remarks on this.

Losing your whole team to one hit and then watching a repeating setup over and over is just bad design.

Solution: There’s no solution that’s going to fix everything. Short of giving characters full invincibility until they touch the ground there’s no possible solution that’s going to get Hulk out of Viper/Strider or Firebrand/Skrull style unblockables. In a game that has fully unblockable strikes, characters without air mobility options are going to be screwed somehow. That’s just life.

In short, Marvel's combos are actively terrible. The game has other features that do make it enjoyable, and it's personally my favorite game to spectate, but I can't stand playing it precisely because of its combo stupidity. The combos are flashy, and difficult to perform, but there's very little of substance present in them. I will grant that it is extremely hype to see creative combos, such as MarlinPie's old Doom TAC back before TAC infinites were a thing. I like that that's possible to do. What I don't like is when one super-powerful option is just so clearly optimal that the character might as well just die as soon as the hit is confirmed, which is what the game is evolving into.

Much better than Marvel is The King of Fighters 13. While not as popular as the Capcom games in the American scene, there are a significant number of people who hold KoF as the most technically-sound fighting game of the current generation. It plays a lot like Street Fighter in neutral, with a heavy focus on ground-based spacing and shorter combos ending in knockdown. But the drive meter system allows you to extend combos for much greater damage, even reaching near-Marvel levels in HD Mode. Importantly, the cost of doing so is very high, and since each player has three characters on their team to work with, deciding when to use it is absolutely critical.

I agree with the aforementioned significant number of people, at least in regards to KoF's combo system. It's my favorite from a theoretical standpoint--but I prefer playing Blazblue because of all the extra versatility and options it affords me.

Combining it all together

The four games I've talked about so far cover pretty much the entire spectrum of combo systems for games I'm familiar with. Melty Blood, Vampire Savior, Persona 4 Arena, and others are all very similar to at least one of them, and the benefits and drawbacks are similar. As far as games I don't know, I'm looking forward to Killer Instinct because of a post I saw discussing how it inverts the usual neutral-combo relationship, where getting the first hit is trivial and almost all of the yomi and decision-making takes place during the combo itself. I don't imagine too many fighting game fans would enjoy that design, but it would probably be fun to look into.

Hopefully, by reading this you attained some further insight into how combos both reflect and shape a game's design. I don't doubt that some of my conclusions are suspect; I'm no fighting game scholar. But regardless of whether or not you agree with my analysis, I hope I've inspired analysis of your own.

I'm not sure what I want my next blog post to cover. I may cover the combo system of another game I'm familiar with--one so vastly different from all of the ones mentioned here that it requires radically different thinking. Or I may talk about the concept of internalization and what it means both in fighting games and out.