Does Chance Belong in Fighting Games?
Updated: Oct 15, 2019
Most popular fighting games are devoid of randomness. It’s a common thought that the more randomness you add to a game, the less skill based the game. Fighting games, usually aiming to be a test of skill, tend to avoid random elements altogether. In the most typical example, Street Fighter, the match begins at the character select screen. At this point, both players have exactly the same options. Other than who is player 1 and who is player 2, it's a perfectly symmetrical game. If both players pick the same character, which is allowed, this symmetry remains right up until the game begins.
This is colloquially called a “mirror match”, and the only reason the symmetry ends here is due to the likelihood that any two players will behave differently, not because of anything the game does. This symmetry is a staple of the genre. Most fighting games operate the same way and do not depend upon random elements whatsoever.
This avoidance of random elements is really an avoidance of unfair play. To illustrate
that, let's look at one of the most popular examples of randomness in what is otherwise an extremely technical, player-driven game: item drops in Super Smash Bros. Items of various effect and utility appear randomly around the stage during a match. Whoever manages to grab the item first, most likely whoever is nearest when it spawns, likely gains some sort of advantage.
At this point I should make clear that I am examining randomness from the competitive point of view. That is to say, how an element of randomness affects a game's viability as a meaningful competition. Many random elements being used in fighting games are much less of an issue when competition is not the end goal and items are a great example of that. They are great fun when playing casually, but are not used whatsoever in competitive Smash. They're turned off at the start. This isn’t because the concept of picking up items and using them to fight is inherently bad, just that due to the implementation in Smash, many items provide an advantage to a competitor essentially at random. The effect of this level of randomness is so significant that, to a competitive Smash player, playing with items on isn’t really “playing the game” at all. Along the same lines, also from the Smash series, is the infamous Super Smash Bros. Brawl Tripping mechanic.
This feature made it so that whenever anyone tried to run or turn around while running there was a 1-1.25% chance that they would trip and fall onto the ground, unable to act for a short time. Recall that Brawl was the long awaited sequel to Super Smash Bros. Melee, possibly the most movement-focused fighting game of all time. Running is something that players want to do often. To be punished for movement, especially at complete random, was nuts. Tripping was so fundamentally against the nature of competitive Smash that it was one of the biggest contributors to competitive Brawl never really taking off at all. The game had other issues, but perhaps if this mechanic had not been such a slap in the face to the competitive community, the game would not be remembered quite as poorly as it is today.
Another interesting example of randomness calling into question the competitive nature of a game is Guilty Gear's Danger Time. Danger Time is a 10-second special state that has a random chance of activating whenever two attacks clash.
During Danger Time, any hits landed on an opponent become hugely damaging combo
starters, much more significant than they would otherwise be. This grants whoever happens to land the next hit a huge advantage for little reason. Unlike items in Smash, Danger Time cannot be turned off. However, Guilty Gear is still played competitively. There is a key attribute of Danger Time that sets it apart from things like Items or Tripping and, while it is still a controversial inclusion, prevent it from destroying the game's competitive viability: Danger Time does not assign advantage randomly. In other words, it does not take away control from the players. The next hit scored does do a considerably higher amount of damage, but it is still up to the players to decide who gets hit. Danger Time doesn’t affect the players’ ability to play whatsoever, just forces them to play at intensely higher stakes for 10 seconds.
A similar example in design, but not in presentation, is the Smash Bros. Pokemon Stadium stage.
This stage transforms over the course of a match into one of four variants. The variants
are chosen randomly and do not inherently affect one player or another. Each variant carries with it different strategies and interesting player interactions, spicing up the competition. In this way, the randomness just affects the circumstances of a match and not directly the players. Even this is sometimes considered a questionable stage to include in competition, but it has been generally accepted for much of Melee’s 15+ years of competition. It is possible that this stage is more well-regarded than Guilty Gear’s Danger Time because there is the extra layer of control provided in at least depending upon the players choosing the stage to be subject to the randomness. Implementing randomness that does not immediately advantage a player is the key to implementing randomness in a fighting game. There are a few examples of fighting games that have explored this idea with varying degrees of success. One of these such examples is Tira from the SoulCalibur series.
Her gimmick is that certain specific moves in her repertoire have the chance to change her into or out of a more powerful state, called her "Gloomy" state. This implementation of randomness is interesting for a couple reasons. Firstly, it puts the player in control of the randomness. It is up to the person controlling Tira whether or not they use the state-changing moves. If their roll of the dice fails them and Tira loses her Gloomy state, the player can be disappointed, yes, but they knew what they were getting into when they initiated the move. The game doesn’t give or take for no reason; it is always a gamble. Players are given incentive to use these moves by making many of Tira’s more damaging combos while in Gloomy depend upon them, at the risk of switching back to her weaker state, called "Jolly". Not only that, but certain combos are forced to change midway through because of a random state change, testing both the player’s reflexes and their knowledge of Tira. This design actually increases the depth of the character and creates a neat risk-reward system baked into her moveset. A secondary effect of these state-changing moves is that the designers are subtly teaching new players how to use Tira. It would be a player's natural inclination to find which moves transition from Jolly to Gloomy and use those moves repeatedly, because Tira is a stronger character in Gloomy. The player's trial and error when using these moves over and over will lead them to an understanding of how and when they should be used more quickly than if they were just looking through a huge move list and using whatever came to mind. Another example of giving specific characters random moves is Faust’s prolific “What Could This Be?” move from Guilty Gear. This is a special move done by the character Faust, in which he reaches into his coat and throws out an item onto the battlefield.
However, the item he pulls out is decided randomly, pulling from a finite list of possible items. Each item is wildly different and behave in different ways. For example, he could pull out a donut that heals whoever picks it up or he could pull out a meteor that waits a set amount of time before descending on the area in front of Faust. Faust’s entire gameplan revolves around the use of this move, forcing both players to react and adapt to whatever it is Faust pulls out. This demands that both players have some sort of understanding of the items and how Faust is likely to capitalize on any given one.
Once again, we see that this use of chance is put in the control of one of the players,
deepening that player’s pool of choices while also requiring the player have a more thorough understanding of the game to be successful. We also see that this move does not assign advantage to either player at random. This is Faust’s special move, and therefore the majority of the items are to his benefit. It may then seem unfair that he can randomly generate items for himself, however, knowing that his character depends upon these random items from the start allows the designers to ensure that they are relatively balanced as a special move. For the most part, his items have a level of utility equivalent to that of a regular special move from any of the other characters. In the same way that they can choose to use their special moves at any point, he can choose to throw out an item. For the more extreme items, varying drop probabilities and regression to the mean work together to prevent them from providing too much imbalance.
These two characters represent some examples of well-received, well-designed utilizations of randomness for fighting games. They avoid the frustration of just handing out a win to a player for no reason by putting the randomness in the control of the players. Rather than working like black-or-white, win-or-lose dice rolls, they reward the player who has a better understanding of the game and can comprehend the results quickly. It is clear when examining the few, rare examples of successfully utilized randomness in the genre that the community favors mechanics that are inside their control. An idea that demands more consideration is if random mechanics existing outside of a player’s control are simply diametrically opposed to the competitive fighting game genre or, perhaps, that these mechanics just have not yet been really explored. This type of mechanic is very rare. The few examples that come to mind are Danger Time and the Pokemon Stadium stage. I hope to see more exploration of these ideas in the future. I don't believe that all fighting games need random elements; they have plenty else besides literal randomness to break up the symmetry. However, when these games do employ randomness, I hope they pay attention to what's come before.