Thursday, August 18, 2016

A Facet of Randomness

Randomness, or RNG, which typically stands for "random number generator" but is typically used synonymously with describing random chance, is an element that crops up frequently in games from basic concepts such as critical hits to more complex ones such as certain AI behavior. It is a potential cause for events both glorious and frustrating, but mostly frustrating. As a result, RNG is discussed often and thankfully not all of them involve rants. What I take away from the frequent discussion (and frustration) related to RNG is that there are ways to design it so that the players involved don't feel cheated somehow. To me, a lot of this comes down to the degree of choice a player has in terms of controlling the RNG.

Based on discussions such as the ones I linked above, I'm apparently not alone on this, so any input I have on the topic will be limited. However, this article has a purpose beyond repeating some points on RNG. Since the pre-patch, I've once again been learning to play Hunter and much like in the past, I have criticism of the class ranging from its thematic to its gameplay. The last time I posted such criticism, I mentioned randomness a few times but the point would've been easy to miss along with the intentions of my redesign since the entire article was comparable to a well-formatted wall of text. Along with addressing the wall of text issue, I wanted to publish this article as something of a prelude to provide context on what I think good and bad RNG are to a degree. To accomplish this, I will describe three (arbitrary) categorizations of randomness with examples mostly from Hearthstone and WoW.

Controllable Randomness

Controllable randomness utilizes random chance, but provides the player with methods to deal with the random chance in hugely substantial ways. In addition, it also helps if the randomness is curbed in other ways such as by reducing the RNG itself. For example, a high chance to activate an effect would be a reduction of RNG since more often than not, the effect in question will activate. Taking advantage of gameplay depth, game knowledge, and the like are strongly preferred bonuses.

Discover in Hearthstone is a good example of controllable randomness because it restricts card choice to specific categories and offers three cards for the player to choose instead of just giving them the card. At the very least, players are given the option to choose the objectively best card of three potentially bad options. However, Discover also allows the player to tune their choice based on situation and game knowledge since the objectively best choice is not always the right one.

In the context of WoW, an ideal example of controllable randomness would be a proc that activates when specific abilities with variant priority of usage critically strike or the like (Chaos Strike, maybe?). This allows the player to feel like their interaction with the game has meaning through their usage of abilities. In addition, the player can build their character in a specific way (in this case stacking crit chance) to increase the proc chance.

Pseudo-randomness

Pseudo-randomness has numerous elements associated with controllable randomness but often fall short in some way that makes the player feel like they have less control over the randomness somehow. Since there's usually numerous facets associated with concepts related to randomness, the pseudo-randomness can vary greatly and is strongly subject to opinion.

The Hunter's Mark passive in WoW is an example of pseudo-randomness since auto shots aren't really controlled by the player, which can make it feel like a proc from this passive comes out of seeming nowhere. However, stacking Haste to increase the number of autoshots increases the number of procs (as indicated by the tooltip itself). This means the player can specifically control one facet of the randomness. It's worth mentioning that this particular mechanic may not seem that exciting since the ways to manipulate the random chance don't involve frequently used abilities, which is a more active style of play compared to building a character (or using a cooldown).

Total Randomness

Total randomness is a type of randomness that is for all intents and purposes, entirely random since players have little to no control over the random elements. It's also worth mentioning that total randomness is associated with a low chance of occurrence, though that's not always the case. In general, such randomness feels like the game is cheating you or that you didn't deserve the outcome due to the aforementioned lack of control. This type of randomness is rarer nowadays and can arguably be confused with pseudo-randomness.

The chance of moves missing in Pet Battles was an example of total randomness. While a lot of moves have been fixed now to have a 100% chance or otherwise emphasize the risk-reward factor with lower accuracy, some moves previously had accuracy values at 80-90% with slightly higher damage values. In general, most pets didn't have abilities to increase this chance to hit so there was a risk of the move missing outright. This is a phenomenon observable in other turn-based combat games like Pokemon or the average JRPG.

Final Statements

As I implied above, these points more reflect my views on the types of randomness that exist (in gaming) nowadays in a rather broad sense that ultimately is determined by the amount of player agency the randomness allows for. While I wouldn't recommend bringing the points brought up here into a typical discussion on RNG since the concept is a little more complex than what degree a player can control it, I would keep these points in mind when I publish the aforementioned article on reworking the Hunter class in Legion in the future. This is because some of the changes will focus on implementing more instances of controllable randomness or improving upon existing instances of pseudo-randomness. I have no idea when the article in question will be published as of now, but I should be able to take some time off experiencing the Legion pre-patch to publish a couple other articles soon, since I have at least one article related to the pre-patch I want to write.

No comments:

Post a Comment