I had the opportunity to attend the Friday events of PAX East 2015 last week, and had a predictably good time mingling with other video game enthusiasts and seeing the offerings of different booths. Most of my time during the day, however, was spent attending the wonderful panels that PAX offers. I wish to offer commentary on two of these panels in particular, which I found especially incisive and provocative. This article is the first of these two commentaries, on GeekNights‘ lecture entitled: “What is Losing?!”
I’d not heard of Rym and Scott of GeekNights prior to PAX East, and I must preface my commentary by saying I was quite impressed by their analytic approach and presentation in general. I found myself agreeing with their arguments more than disagreeing; however, there were points at which our views diverged. In what follows, I hope to offer readers a critical perspective on what about the view’s presented in the lecture were tenable, and what might need refinement. My intention is that such commentary will be useful in directing future inquiry into advancing the analytic tools of game analysis.
First, I’ll offer my gloss of the major arguments made by GeekNights in their talk. I should point out two things: firstly, their PAX East lecture, as far as I can tell, is not yet online, but you can find an excerpt of a similar talk which they gave at PAX Australia 2014 here (I will update this article if their PAX East talk becomes available online). Secondly, their talk covered a lot of ground with regards to the domain of gaming, from single-player video games to multiplayer video games to board games to tabletop games. This means that my own analytic approach of dealing primarily with single-player RPG video games is going to end up bracketing a lot of their talk; however, I think we’ll find there’s still plenty of material left for us to dig into. (I grant, of course, that I could be reading GeekNights differently than they intended at certain points in their talk; my aim is to represent their argument as faithfully and charitably as possible after having seen their presentation.)
Here’s my gloss of GeekNights’ main arguments: they’re pushing back against the assumption that the goal of games is always to win. To do so, they offer a variety of type-wise distinctions between the concepts of losing in various games. These are the basic forms of losing they pick out.
Losing as an obstacle: this comes closest to the “standard interpretation” of losing. Under this view, losing is merely an impediment to eventually winning the game — think, to use GeekNights’ example, of death in “Super Mario.” The only narratological significance losing has here is to effect winning, which is to say that the player cannot logically win a game unless it is also possible to lose.
Losing as masochism: terrible though this term may sound, think about a game like “Five Nights at Freddy’s.” The argument is that ‘losing’ in this game is deliberately implementing losing as a mechanism for horror: by temporally separating the moment when a player “loses” the game and actually dies, the player is left unable to anticipate the game, an experience which is unsettling because the player cannot use loss to quickly understand the causal relations between her actions and the feedback within the game’s world. This kind of masochism, perhaps better glossed just as ‘horror’, is a unique thematic effect made possible by the dynamics of losing.
Losing as punishment: this form of loss maps onto “ruthless” games, like “Dark Souls.” The idea is that losing creates difficulty and frustration that makes eventual success more pleasurable — so in this view, the dichotomy of loss and victory reduces to hedonic interests of pain and pleasure.
Losing as a directional narrative force: GeekNights talks about games whose plots are “pushed” by failure — in other words, failure merely pushes the storyline in another direction. The archetypal example offered is “Chrono Trigger” (Square, 1995), in which “success” or “failure” in certain story events branches the narrative in different directions, ultimately leading to thirteen different endings. So rather than failure stopping the story and forcing the player to try again, this model of losing actually requires that the player fail at certain points if she wishes to pursue all possible outcomes of the story.
GeekNights abstracts from these definitions to construct an argument against what they call the ‘narrative of victory’; they claim that good games actually implement ‘narratives of failure’. They first offer the example of “Super Meat Boy,” where completion of the game is “celebrated” by a montage of every failed attempt leading to ultimate success; then, they abstract to games like “The Stanley Parable,” in which no definitive plot is privileged, but rather a diversity of unique exploratory paths through the game is offered. They actually coin a term to define this modality as a particular type of game: they call it an “idiogame’, “a game presenting a series of interesting player decisions leading to a personal outcome.” Their key takeaways are fourfold: “losing effects winning,” which is to say that winning and losing are conceptually codependent; “losing is contextual,” which is to say that individual games use the concept of losing for different narrative purposes; “losing is optional,” meaning that some games don’t implement the idea of losing at all; and, understandably, that “games should be fun.”
So much for my rendering of their argument. As I said at the outset, there is much that I like about GeekNights’ talk. If you’ve read my article on “Dark Souls” (and you really should), then you know I think it obvious that games can implement death and failure in ways more significant than mere obstacles to success — e.g., in the case of “Dark Souls,” the opportunity for transcendence on the part of the player. And GeekNights does seem to pick out important differences in the types of losing which video games can implement. However, there are two issues on which I would push back their analysis: one is the conceptual boundaries between types of losing; the other is the status of idiogames relative to other types of single-player RPG’s.
It seems self-evident to me that games can implement losing, like difficulty more generally, in different ways. However, if we are to taxonomize these differences, we must do so in such a way that the resulting taxonomy is accurate and analytically useful. While the classification offered above is a good start, I think the model suffers from issues of taxonomic overlap and vagueness.
Here’s one example. GeekNights distinguishes losing in a game like “Super Mario” from losing in text-based choose-your-own-adventure games (old school computer games) in the following way: for “Super Mario,” they claim, there is no interesting, unique story told based on Mario losing by falling into, say, a particular pit in a particular level; on the other hand, the choose-your-own-adventure games in question often have vivid narrations of how you die when you lose in particular ways, which makes it interesting to explore dying in different ways. But the “Super Mario” example seems like a straw man to me; when you instead consider games like “Legend of Zelda,” I think it becomes apparent that this flavor of losing is just as conceptually interesting as the choose-your-own-adventure case — otherwise, we wouldn’t have montages like this. What GeekNights is picking out here, I would argue, is not a difference in concept so much as a difference in medium: what text-based games represent in written narrative, more modern video games represent through graphics, sounds, &co. By stipulation that characters can die in different places across the course of a narrative, it follows that their deaths will not all be identical, even if the degree of difference feels more superficial in the case of “Super Mario” than in the case of “Legend of Zelda.” So maybe we will want to say the a game like “Super Mario” doesn’t deploy this concept in an aesthetically interesting way, but such analysis depends on us first establishing that these different media are deploying the same concept of losing — i.e., losing as an interesting obstacle that changes with the ecology of the game and the player’s choices. I think such an analysis is more parsimonious and will allow us to better analyze games in a comparative framework, because we will be working from a common conceptual base.
Another example, which I think generally requires more research, is the question of whether ‘losing as an obstacle’ and ‘losing as punishment’ are conceptually distinct. In all honesty, it wasn’t clear to me whether or not GeekNights was claiming that they were — though I would say that they probably were. The issue for me is where difficulty enters into the equation. Player skill almost always factors into the frequency of losing within a game; so a game like “Super Mario,” to a first-time player of video games, might be as challenging as “Dark Souls” is to a more experienced player. This would suggest an interpretation by which the ‘punishment’ concept, as stipulated above, might be a difference in degree from the ‘obstacle’ concept, rather than a difference in kind. Where this becomes interesting is in analyses of the differences between the following categories: games which allow you to pick a difficulty level, but require you to keep that difficulty level through the entire game (“Kingdom Hearts”); games that allow you to change the difficulty level at any point (“Dishonored”); and games that don’t have difficulty settings (“Dark Souls”). It seems like we would need a particularly robust analysis here, integrating the skill level and aesthetic desires of the player. So the concept needs to be explored further. (It also goes without saying that, on such an analysis, we would need a distinct concept of losing for games like “Dark Souls,” where dying, though certainly an obstacle, facilitates unique ways of the player experientially progressing, in terms of despair and transcendence in the face of nihilism.)
The second issue, which I find much more interesting, is the question of just what an ‘idiogame’ is. The stipulation, as I take it, is that certain games deny the notion of what I have been calling a main plotline, allowing the player to merely adventure through the world presented to them, thereby crafting their own story. We can imagine the implications of this by recalling the model I’ve advanced of video games being represented by a narratological three-space:
For a full explanation of the model, refer to my paper, Player Agency in Majora’s Mask. But suffice it to say as a refresher that the z axis describes a game’s main plot from its beginning at z=0 to its end at the maximum z value; the (x, y) plane, in contrast, represents all optional, ‘exploratory’ paths through the game available to the player. A game’s narrative in a given playthrough is described by the sum of vectors taken by the player from z=0 to the maximum z value.
Now, idiogames might be able to be schematized within this framework as follows: if, by stipulation, there is no main plot, then the game will instead extend only as an (x, y) plane, with players able to explore the world of the game as much as they wished, never encountering a set ‘end’ to the story; instead, their playthrough may merely terminate when they die, or something like that. This does indeed retain the exploratory and personal element that the definition of idiogames has in mind; the problem is that it’s hard to see how the formalization defends an understanding of a given “exploration” as a discrete story told within the world of the game, because we have no main plot to refer to as a reference point. Such a model feels instead much more like what is commonly called “sandboxing” — think of “Minecraft” or a big bin of Legos to build at your discretion — in which the player can explore the world presented by the game with very few limitations, but in which the elements of narrative are few and far between. We might want to bite the bullet and say that, in such games, the player is doing something more like “brainstorming,” actively creating a story instead of engaging with and interpreting a story that was mapped out prior to the player encountering the game; but if we want to hold on to our notion of the game as a narrative object, then this seems insufficient because it describes the video game as a notebook in which to write ideas, rather than a novel to engage with in dynamic ways. So the idiogame seems like a fascinating notion for which the analyst ultimately must answer; but it’s not clear that, as presently defined, it escapes the conceptual dilemma presented above. We will need more analysis to resolve the tension, perhaps either by defining them as something different from a story-driven video game, or by developing a more intricate schema for relating it to our narratological formalization.
Again, I found GeekNights’ lecture to be wonderfully entertaining and insightful. But it also ought to serve as an indicator of where we need to dig in analytically to more rigorously define our terms and parse out the narrative mechanics of games. Resolving the issues presented above would be a significant step on the road to cleaner, rigorous analysis of video games.