Words: 2420 Approximate Reading Time: 15-20 minutes
It’s possible that a fair number of people were made aware of a new announcement this week out of Activision Blizzard. On Thursday the company issued a press release about a new piece of internal software they were rolling out called the King Diversity Space Tool. The proposed objective of the tool is to provide a way of seeing how diverse particular characters within a game are – and all characters in a game as a whole. The idea is to foster better representation within video games.
Let’s back up for a moment. There has long been known to be a representation problem within video games. If you were to pick a major title at random, you would have an overwhelming chance of picking a game with a main character that was white, cis-male, and straight. Basically, a massive number of main characters sort of…look the exact same.
Now this is, in fact, a problem. It is highly limiting from many perspectives. Narratives essentially have to explore the world through the perspective of the main character, and while you can explore a large number of broadly human stories, the ability to engage in social or political commentary is constrained by the type of character you are saddled with.
But also important is that it limits representation in video games. If nearly every main character looks the same, then the kinds of people who will want to play those games will be those who see themselves reflected in the main character. The ability for us to see ourselves represented in the world in various ways – for minorities of all stripes to feel recognized by seeing someone “like them” in art and culture – is an important component of expanding video games in general. The more people who are playing, the greater appeal, the more people making games, the more there is to consume, and so on and so on.
There is a lot more that can be said about this subject. The above few paragraphs have barely even scratched the surface on the topic of representation in video games. But there has been so much written on this topic that I could really on retread arguments made by others. So we will lay down what has been said here as sufficient. But here are a couple of articles on the issue if you want to do some further reading.
The focus here is on that diversity tool that Activision Blizzard has announced. Because…it’s terrible. If you’ve heard anything about it, you’ve likely seen people talking about how the tool is basically asinine and doesn’t really solve the problem it is supposed to solve – in fact, it likely is counterproductive.
So I wanted to explore why people are basically laughing at this new tool. And more broadly, the topic of implementing diversity in video games and how a “tool” as being discussed is ultimately not the solution – which is something that those who have really looked into the philosophy of science more broadly would immediately be aware of.
Representation and Tokenism
So if the problem is that there essentially aren’t “enough” people who aren’t white, male, and straight in games, then surely that means all you need to do is just add in various non-white, non-male, and non-straight characters, and you’ve accomplished your goal, right?
Well, no. This brings us to the problem of “tokenism.” The term refers to the way in which a person – particularly someone from a minority group or culture – might be inserted and/or highlighted in order to make an outward claim about whoever is engaging in the tokenism. So for example, a business is getting complaints that its workforce is too white, and so it hires a single non-white employee, and points to that employee as proof that their employees are “diverse.”
Tokenism is a form of deception, by way of deflection. It is a way to escape criticism by pretending to care about issues of diversity, without investing any real care or effort into solving the problem. Hence why the effort is so minimal, but the rhetorical aspect is so loud – we hire a single employee, but put that employee front and center and announce to everyone that we’ve made this diverse hire. The point is not to hire that employee because they have skills suited to the job and therefore contribute to the business. They are hired because the management team wants to avoid being called out.
The problem with talking about tokenism and pointing it out in more concrete forms is that it invites various bad faith responses. If, for example, you attempt to call out someone for engaging in tokenism, the answer is “what, you simply want me to hire more non-white persons?” While the answer may in part be “yes,” we still don’t get to escape the underlying problem.
Which brings us to the nature of representation. While mere visual representation is good, the tendency for such representation to devolve into tokenism is not only understood, but is something that consumers tend to see through pretty easily.
Which means representation needs to be substantive. Again, there’s a lot of discussion surrounding this topic, but to simplify, you want representation to be done in a way that focuses on the individuality of the person or their experience, rather than on their mere characteristics. If you were including a non-white character into a game, it’s not enough that they just “look different.” You want the character to feel authentic.
Generally, the best way to accomplish this is through incorporating perspectives of these people in the process. When it comes to hiring a more diverse workforce, you start with bringing in people into the upper echelons so that they can help determine the best ways to hire a diverse workforce. If you’re including minority characters into a story, having people on the teams that can speak to what life is like for someone in that group helps to make the character feel more real.
Which is why the solution, as so many commenters have been pointing out, is not this tool, but seeking out more employees from diverse backgrounds. It is about forging a work culture that encourages people who fall outside the bounds of “normal” (i.e. white, male, straight, cisgendered, neurotypical) to stay on board and voice their ideas.
That this tool is being revealed at Activision Blizzard, which is under scrutiny for fostering a toxic and harassing workplace, helps illustrate the basic problem. Good representation ultimately needs to start with the employees themselves, and if employees are quitting because they face various forms of harassment, then their perspectives get lost, which in turn leads to bad representation in the games themselves.
Philosophy and the Weakness of Technology
So why can’t this special tool solve the problem?
To begin, let’s go over what we know about the tool itself.
The tool is a way of quantifying the diversity of a given character in a game. By designating certain characteristics (race/ethnicity, sexual orientation, gender identity, etc.) as “more” and “less” diverse, you can come up with a sort of space to determine how diverse a character is. The more points on the more axes, the more diverse they are. The premise of the images that were initially shared (before they were taken down) was that the measurement was on the “space” created by these axes, rather than just quantifying a single axis (hence, almost certainly, the name of the tool).
Unfortunately, at this point Activision Blizzard has edited the initial entry, which originally contained various images showing what the tool looked like in practice. Thankfully, several users on Twitter shared and saved images that we can look at. From those images, we can see the list of characteristics that they use for measuring diversity:
- Cognitive Ability
- Physical Ability
- Body Type
- Facial Features/Beauty
- Gender Identity
- Sexual Orientation
- Socioeconomic Background
Each entry – at least from what we’ve seen from the outside – is then given a particular score based on (apparently) a drop down menu. If a character is of such-and-such race or has such-and-such physical impairment, each of those is worth a certain number of points on each individual axis.
The problem with this is twofold.
Firstly, it essentially reduces all characteristics to a set of numbers, which in turn assigns value to them all. The implication of these numbers is that on each axis, particular forms of diversity are worth more than others.
So take the particular example from the above tweet: the character Ana from Overwatch. Ana’s culture is Egyptian, which has a score of 7. Intuitively, that makes a degree of sense – Egypt is a part of the world that is certainly underrepresented within media (at least beyond portrayals of Ancient Egypt). But then by giving it a score of 7, that means it is worth less then whatever might score an 8, a 9, or a 10. And conversely, a character from a country with a score of 5 or 6 is worth less than Ana – at least on that axis.
It is the inherent problem with quantification itself. While we love having data and talking about the world in terms of numbers – since numbers are ultimately easy to grasp and feel objective – this data ultimately does not tell the entire story. It doesn’t really tell a good story. But because it’s captured in numbers, and bigger numbers are better, we go with it anyway and don’t bother inquiring further.
It must also be noted that from the images we saw, it appears that the tool only allows for one trait per axis. So if a character has more than one physical impairment, or has a background in more than one culture, or anything like that, then they appear to only get one entry. Perhaps there is a “multi” option within these categories, but even then once you introduce quantifying these concepts, you necessarily introduce the problem of figuring out how all of these “multis” compare to each other.
Secondly, it still tokenizes representation. In fact, by quantifying it in this way, it is encouraging tokenization. Because remember, the nature of tokenization is about not just having representation, but that representation existing for the purpose of showing off that representation. And so the premise behind the tool itself is creating sufficiently diverse characters not because the design team actually wants to have characters with those traits, but to avoid critiques that there isn’t “enough” representation in these games. It takes us back to the bad faith responses.
How do we know that it is encouraging tokenization? Because ultimately, the only way to get an overall better score on this diversity spectrum is to start with an important question: “how can we make this character diverse?”
It’s a subtle question. It’s easy to see it as innocuous. But the outcome is that you’re still starting from the position that there are characters that are “normal,” and some of them need to be “de-normalized.” Rather than seeing all of these different characteristics as normal and including them naturally, the inclusion is inherently artificial. As opposed to when we have a diverse group of people making the game, who can then make these kinds of suggestions naturally, because they can immediately see problems and address them.
The introduction of this tool ultimately speaks to a broader problem with society’s fixation on technology. Whether it’s the use of AI in a variety of contexts, blockchain, or statistics, we’ve become more and more enamored with the idea that if we just quantify every problem or turn the thinking over to computers, we will have solved all of the issues that human beings face.
And that solution is just…wrong. Whether it’s the inherent racism and sexism built into many AI programs, or the way that blockchain technologies create a cult of technological worship around the resulting cryptocurrencies, or the use of bad statistics in all sorts of ways, technology ultimately reflects our own problems as humans.
Which is why this tool is so…horrifying? Because even if you layer in a bunch of stuff on top of it, you’re still starting from a rotten foundation – that if we just quantify the diversity of characters and get bigger numbers, that’s better.
And again, this is why so many people who have been going after Activision Blizzard for this reveal – delivering criticisms which ultimately led the company to make significant changes to their post (including removing all of the pictures of what it looks like). Because the solution is figuring out how you make the inclusion of these characters natural. The characters should not be introduced because you have to include them to check off some imaginary “Diversity” box. Rather, they should be included because you have a member of the team saying “hey, I’m non-binary, and maybe having a non-binary character would be a good idea.” You get the same ultimate outcome of a more diverse cast of characters, but the basis for how it happens radically changes.
If there’s one thing that studying philosophy helps you to do, it’s to see the underlying faults in our thought patterns. We as humans have tons of bad habits, such as biases, blind spots, a desire for shortcuts, and so on. Philosophy certainly isn’t the only area of study that allows you to see these problems, but it certainly helps to highlight how prevalent they are.
And the hyperfixation on technological solutions to these underlying biases becomes another bad habit in its own right. The King Diversity Spectrum assumes that the problem of diversity in video games is entirely a problem of diversity within the games themselves. But the problem is actually a problem of the workforce. Game development, especially at large studios, faces a diversity issue, one that has been known for a long time. A problem about not just employing a diverse group of people, but also putting those people into higher positions and treating them with respect and listening to their ideas.
The true solution to the diversity problem within video games is addressing the problem outside of video games. Things like this tool are at best only half-measures, and even then not very good ones. A tool is only as good as the purpose it is ultimately used for. This tool in question serves not to actually foster substantive diversity within video games, but to address a core criticism within gaming culture by papering over the issue.