From Animal Crossing to Battlefield, randomness is a huge part of our medium of games. Random number generators decide which Pokémon we find in the tall grass, what items fall out of our enemies, how much damage we deal, how our bullets spread, how we grow and level up, and even how AI behaves. This randomness rules our virtual lives in subtle and not-so-subtle ways. It creates uncertainty, which makes things more interesting, but sometimes that uncertainty is unwelcome. Even with all the randomness of games and the fun of uncertainty, there is a certain amount of consistency that I like to see as a player. Without it, the decisions players make become less significant, and the game weaker as a whole.
To understand why the consistency is important, it helps to know why the randomness is, too. The shortest version is: not knowing is exciting, down to a deep psychological level. Those familiar with B. F. Skinner or fans of Extra Credits may know that adding variation to reinforcement makes things more engaging. This means that, when you throw a Poke ball, for example, it succeeds every time, the Pokémon games would be a lot less interesting than if it had a chance to fail (and that chance changes based on health, status condition, catch rate…). The foreknowledge that throwing a Poke ball will not always work is what makes catching Pokémon feel rewarding. The fact that the ball can fail to catch, even against weakened Pokémon, is part of the genius of the series. Another good example by the big N is Animal Crossing, in which pretty much everything from day-to-day is random: the shop items, your neighbors’ moods, what you dig up or shoot out of the sky, whether your flowers have hybridized yet, but while those things are a big part of the game, they are not all that influential on the player’s decisions. How one decorates their house or town or which neighbors they like to pester the most, while potentially sped up or slowed down by the randomness, is not otherwise reliant on it. The whole game is about taking it slow, so it does not matter how long it takes to fill your house with blue roses placed on custom-designed tables. The fun of the game is the journey.
Of course, this comes with a tradeoff, as all good things do. For starters, Skinner-box methods are a cheap and shallow method of adding engagement. Players will probably not get bored immediately because the excitement of “what will happen this time?” is enough for a little while, but it is not interesting enough to make them genuinely care. Pokémon, to bring up a previous good example, makes it work because it makes you work for that chance at success. Even outside the context of catching ‘em all, or even just catching that one that you like, the game sets up a situation in which randomness plays a part, but does not entirely decide the player’s outcome. However, I do not want to imply Pokémon has perfect use of RNGs. Occasionally it does tip the scale too far in favor of the randomness, and catching a Pokémon becomes next to impossible without luck even with all the right conditions (I’m looking at you, Articuno). This domination of randomness is also why competitive battlers often ban evasion and accuracy-altering moves and abilities, as all they do is add a miss chance decided by an RNG. Now this sort of randomness is a method of creating difficulty using systems already in place. The trouble is, whether it aims to create difficulty or not, it is always frustrating.
Using that Skinner model again, in theory the less likely something is to work when given an input, the more one will want to try giving an input–this is the very foundation of slot machines and lotteries. However, when games put too much emphasis on the random, either in an attempt to create engagement or simulate difficulty, they tend only to succeed in being exceedingly annoying. For one, complete randomness is boring, which is never something to strive for in a medium of entertainment. After the initial thrill of uncertainty, random chance becomes uninteresting as there is nothing players can do with it other than hope. perhaps the bigger issue, though, is that randomness weakens the value of or completely removes a player’s choices.
Games, a medium entirely built around the fact that its audience can interact with it and make their own choices when playing, cannot stand without player choice. As important as the uncertainty aspect is–it would not be much of a choice if you knew the outcome beforehand–there must also be something concrete to base the decision on. The player must have a certain amount of information they can be certain of, or at least believe they are certain on, to base his or her decisions on. If someone knows getting hit will deal them damage, they now have to make the choice whether or not to avoid getting hit, or better yet how to avoid getting hit. If a player knows a character hates chocolate, the player then has the choice available to give her something else or give her chocolate anyway to spite her. The player remains relatively uncertain of the outcome, but has something to base their decision on. If these two examples were good examples, the player would have even more to go on, such as a tell before an attack lands or a previous experience with the character.
These kinds of uncertainty are great because they give the player a chance to consider the result of their actions. If he or she dodges left, damage might be avoided, if he or she gives chocolate, he or she will probably disappoint the character: these are conclusions players can come to on their own using what information they have, which makes their choices more meaningful. The player still does not know for sure, but can stipulate. A weakness of random uncertainties is that it does not have this trait. The sheer nature of randomness means there can be no real thought process to aid in the choice made, which makes the player’s role much less significant.
A while back, the game League of Legends had dodge mechanics based on a miss chance. Riot Games did away with it, however, because the developers and players felt it added too much luck to the equation. Stacking evasion was a choice to be made, but even with a ton of it all it would take is one lucky shot, decided not by a player but by the computer, to turn the tides of a fight. It felt cheap and was not all that satisfying for either side, and the game is stronger now without it. In fact, League of Legends has done away with any randomness in its mechanics. Since the whole game is built around team player-vs-player games, the uncertainty is already there: other people are the most unpredictable source of it possible, but they are not truly random. People can be watched, read, and outplayed, which is a solid source of uncertainty. There are countless factors from which players can make their decisions, which in turn affect the decisions of everyone else playing, making a perfect feedback loop while never actually doing away with the uncertainty. League in particular does a good job of it, as while the game can be quite snowball-ey, it does have a negative feedback loop built in through player gold rewards, which means even a team that is behind can turn things around if they are clever about it. Even the team ahead is never truly certain in their victory, which makes it continue to be interesting for all parties involved.
Randomness is not an inherently bad thing. The wastes of Borderlands would not be as much fun without the randomness of its prize drops, nor would rougelikes with procedurally generated worlds, yet random number generators are not the only way to create engagement through uncertainty. What is more, without a certain amount of certainty to balance it out, too much unknown makes for a frustrating experience. Consistency and stability are as important a part of game design as uncertainty is. Without it, there are no mechanics to master or tricks to learn, and no value in our actions as players. Both sides of the equation are important, how else would I fight to weaken my opponent and still hold my breath when I throw an Ultra ball?