Monday, December 23, 2019

That thesis I wrote about speedrunning

In 2018, I wrote a sociology master thesis about speedrunning. Here is a summary of it.

The impetus to write about speedrunning was born after watching an AGDQ video in conjunction with reading one of Weber's warnings about how instrumental rationality will destroy the human soul, turning us all into uncaring machines dutifully performing our duties with swift efficiency and as little sentimentality as humanly possible. Weber cautioned that this future world would be akin to living in an iron cage, where all the bars were made in accordance with all the best state of the art practices, but with little thought of the consequences of building such an inescapable cage. When all you have to care about is the bottom line, nothing else matters.

Speedrunners, on their part, have exactly one goal in mind: to finish a game as fast as possible, using every and all means available to them, up to, including and going way beyond breaking said games in the process. What some call game breaking glitches, others call time skips. Gotta go fast, after all.

To an untrained eye, the parallels between an unfettered commitment to maximize efficiency at all costs, and an unfettered commitment to minimize time at all costs, might seem obvious. However. There is this thing in the academy where obvious things have to be explained in roundabout ways in order to be considered obvious. Gotta go slow in order to explain why you gotta go fast.

This presented something of a challenge in my research review process. As you might imagine, there is not a great body of scientific literature out there on the topic of speedrunning readily available, so I had to find a proxy for it. A first foray was to find out what others had written about Weber and instrumental rationality, to find a continuity of thought to draw upon. The results of this first foray was a scattershot of articles who all pointed in various directions, all interesting in their own right but not too useful when it comes to confidently making statements such as "the current research says x". "The current research", in whatever field you might be interested in, generally consists of more than one or two interesting articles.

Thus, I operationalized the drive towards maximum efficiency in narrowly specified metrics into a current trend to do just that in the academy, commonly discussed under the rubric of "audit culture". (To be sure, said process takes place in other institutions as well, but seeing as it is academics who write academic articles, the academy gets overrepresented in the literature.) In reviewing the articles on audit culture, I found what I was looking for: a broadly coherent set of observations and arguments all pointing in the same direction, most of them lamenting that the increased pressure to perform according to arbitrary impact ratings and bibliometric rankings stifled the lifeblood and creativity of the academy as an institution. Finally, I had sure scientific footing when talking about modern incarnations of Weber's iron cage; the bars are in this cage not made of iron, but rather of the slightly less well crafted slogan "publish or perish".

This presented me with the problem of how to summarize Weber in an efficient manner. Those of you who have read Weber know that he was a bit of a speedrunner himself, laying out his writing as a set of logically coherent propositions which lead to a general conclusion. Upon having presented this conclusion, he then expects the reader to tease out all the social and economic implications by themselves, whilst he goes off laying out another set of logically coherent propositions. This makes for a speedy read, since the relevant section rarely spans more than a couple of pages at most, but it is not advisable for a master student to simply quote a source and say "he has spoken". Gotta achieve the word count, after all.

Fortunately, others have tackled the same problems as Weber, and a later author who dealt with the same problem with Weber firmly in mind was Habermas. Habermas, in his decidedly unspeedrunny thousand page tome Communicative Action, pointed out that a great source of distress for Weber could be sourced to a grammatical unclarity. In short, Weber didn't distinguish between subjects, and thus jumped willy-nilly between speaking about the rationality of individuals and the rationality of institutions. This is a bit of a problem, since it meant that when Weber found that instrumental rationality outcompeted all other values in institutions, it led him to think that said other values went away in individuals as well, and vice versa. This led to a dystopic vision where instrumental rationality as a disembodied whole held society in an iron grip, from which nothing could conceptually escape.

This conceptual confusion gave me an opening to ask just who the rational subject is, and what they are trying to achieve when they go about being rational. Drawing from my other master thesis, I posited that rationality could be something that happened on a group or community level, rather than in a single individual or in society as a whole. A group such as - to take an example completely at random - the community of speedrunners.

Here, I should add that Weber distinguishes between four different kinds of rationality. Instrumental rationality, as you might have surmised, is concerned with achieving a specific goal, and being able to continue to achieve it sustainably for an extended period of time (Pheidippides, the original marathon runner, does not qualify as being instrumentally rational; while he did manage to deliver the message, falling down dead after completing his task makes it difficult to tick off the sustainability box). Weber also uses the concept of value instrumentality, i.e. something being done because it is the right thing to do. The exact nature of "right" varies from time to time, but we can count aesthetics, justice and honoring the dead to this category. To exemplify the difference: while putting flowers on someone's grave might not attain a goal in terms of instrumental rationality, it scores high marks on the value rationality scale. The other two kinds of rationality Weber discusses are tradition and strong (albeit temporary) emotions; these didn't come into play, so I will leave them with an honorary mention.

Back to our beloved speedrunners. The way I examined them - gained access to the empirical data, in science speak - was to watch AGDQ videos. Usually when trying to examine human actors acting in specific circumstances, the scientist is required to undergo a lengthy process wherein they get to know the people involved, the situation these people are in, and the proper questions to ask to get the kinds of answers that suit the research aims. That is, if said people even tolerate the scientist poking their nose in where it does not belong long enough to allow for questions of any kind to be asked. The great virtue of AGDQ is that the speedrunners are explaining their thinking to a general audience, along with a hands-on demonstration of what they do in the most practical terms possible. In short, getting access to the empirical data was a straightforward and speedy process.

Of course, a single person can not capture all the complexities of the speedrunning community in one single thesis, so I had to limit my scope somewhat. I chose to focus on three games in particular: Prey (2017), Diablo 2 (2000) and The Legend of Zelda: Ocarina of Time (1998). These were chosen in order to demonstrate three different important strategic considerations of going fast: glitching through walls & sequence breaking (Prey), RNG manipulation (Diablo 2) and route planning (Zelda). I am happy to report that all the strategies I so carefully outline in my thesis are hopelessly out of date at this point, and that going for flashy quotes rather than descriptive depth was indeed the correct way to go.

When writing this section, I faced the problem of how to present said strategies. The academy is not known for its detailed understanding of electronic entertainment, and merely gesturing to things being well established tropes would not cut any amount of mustard whatsoever. In the end, I fell back on the Weberian habit of presenting an ideal type in order to have something to compare actual reality to. Which is to say, I presented readers with an ideal type casual player, and the game loop said casual would likely fall into during their gameplay. This could then be contrasted to the, ahem, chairs to the walls shenanigans undertaken by the speedrunners.

The conclusions drawn from this study are somewhat counterintuitive, and I find myself at something of a loss in how to best convey them. Fortunately, as the sole author of my thesis, I have the privilege of being able to speedily give myself permission to quote extensively from the source material, so here is most of page 35 and a respectable part of page 36:

A defining characteristic of Weber’s ideal types of rationality is that they are not bound to any substantive defined content. As ideal – almost platonic – forms, they are informed by the context within which they find themselves, and the actors performing meaningful actions within these contexts. By situating rationality in the community, as per Wenger (1998) we can overcome the contradiction inherent in Weberian rationality pointed out by Habermas (1987). There is an aesthetic, value-oriented rationality employed by what I have thus far referred to as the ideal type of ordinary gamers, and another value-oriented rationality employed by speedrunners. Neither of these are wrong or – worse – irrational, but they have to be understood as different methods of attaining the same type of goal defined by the same type of rationality. The fact that these approaches have a tendency to be mutually contradictory when compared across communities does not take away from their rationality – it merely underscores the situational nature of rationality as a concept.

 More importantly, this allows us to sidestep the trap of relativism. There is not one kind of rationality in one community, and another in another, where the epistemological rules of the universe are different and incommensurable. Nor is there a single defined rational course of action, where all other alternatives are irrational. Rather, we have a framework from which we can refer to different practices with commensurate results; different communities try to attain the same goals, albeit with different means. By bringing this to light, we have gained useful information with regards to the object of our study: the speedrunner’s relentless application of instrumental rationality does not undermine the gains they make in terms of aesthetics, but rather constitute it. Completing a twenty hour game in seven minutes is a beautiful thing, as is the realization of an optimal pathing solution in Ocarina of Time or the improbable conjecture of unlikely events of a Diablo 2 run where everything randomly goes as planned. The fact that this beauty may or may not be scrutable from an uninitiated perspective is beside the point.

The thesis then goes on to yell at academia for falling prey (pun intended) to the tendency to chase after arbitrary metrics, and yelling some more that the speedrunners might make a better job of it by making it an all out aesthetic virtue. It is, after all, better to strive for virtue than to grudgingly accept an arbitrary activity as an all-encompassing, soul-draining sidequest which distorts the whole purpose of the institution you find yourself in.

I will conclude with another long quotation (from pages 37-38), not just because it allows me to write this post quickly (which is another intended pun; as is the reference to Keenoy in relation to wall glitching), but also because it underscores just how much insight into contemporary social processes remain to be gleaned from the world of computer games and online cultures. The internet stopped being new a generation ago, and we have to drag academia into this new age, kicking and screaming if necessary. The coming decade will be very interesting in this regard, and I suspect that in some cases, we gotta go fast.

The lesson to take home from this is not that the speedrunners have the right idea, in that we as academics should abandon all hope and embrace the metric. Rather, it serves as an ideal type against which to compare and contrast our theory and practice. It is possible to perform the metric to the exclusion of all other things, thus accomplishing the academic equivalent of finishing a twenty hour game in seven minutes. It is also possible to insist that something will be lost by adopting such a methodology, and that there is an inherent value in taking the slower, more comprehensive route. Sometimes, knowing how the former can be performed can save time and generate useful insight into the various things we take for granted (walls are, after all, very substantial guidelines [Keenoy 2005]); at other times, knowing when the latter is appropriate leads to approaching the long haul with better preparation and readiness. The essence of autonomy is to not automatically choosing the one or the other, but to be able to make up your own damn mind about what to do and how to go about it – and why.

The humanities, in particular, are at the center of this problematic. On the one hand, they are being asked to justify themselves according to metrics that are very far removed from the autonomous values they hold and educate; there is a very clear gap between long-held values and institutional demands. On the other hand, those departments that manage to secure funding with reference to the useful skills gained by delving deep into our shared history – do so only by in an ever so gradual way renouncing these long-held values so as to approach the metric. In an era of declining funding, accepting heteronomy might be the only way to remain within the walls of academia. Again, the lesson is not to automatically go for either maximum autonomy or maximum heteronomy, but to think through the situation and face the fact that there is a choice to be made. Moreover, it is not clear whether performing the metrics or asserting the traditional values inherent to the humanities is the right course of action; ‘right’ and ‘rational’ are not necessarily synonymous.