Reconstructing Race and Ethnicity in Covet Fashion

Here is a thorny question: how can you get people to live up to their stated ideals? Specifically, in this case, the consensus that the beauty of women of all ethnicities should be respected?

This is a question we may ask in many contexts, but I'll address it by looking at the game Covet Fashion. This is a social, app-based game that is very popular, with 2.5 million active users around the globe. As someone who studies sociology of the body and gender, I became interested in the game because its player base is mostly women and girls--in contrast to the male domination of MMOs or gender balance of virtual worlds that I have researched in the past. While some popular MMOs only let you play as men, in Covet, players compete by styling only women models. The gender dynamics are indeed interesting--but my attention was quickly drawn to issues of race and color in the game.

The Problem

I've been playing Covet for about a year now, and I quickly became interested how race and ethnicity play out in competitions. This has been a controversial issue in Covet for some time. The way the game works is that players are given a photo of a real life model and one-paragraph description about her (the inspiration or "inspo"), and compete in styling their game character (or "doll") to embody this theme in an appealing way. The inspo photos are fairly ethnically diverse. Players not only dress and accessorize their "doll," but select her skin tone, features, and hairstyle. Thus, it's possible for players to style their game model to approximate the inspo model, not just in outfit, but in ethnic features and skin color, if they wish. After a period of time, the window for players to submit their looks closes, and players then vote on a series of paired looks. Players who do well in a competition receive prizes of clothing and game currency. And those with perfect scores in voting who have also used only in-season items to create their look, and have used a lot of unworn items, receive "Top Looks." Receiving a Top Look gets a player a substantial prize of in-game currency, and their doll is displayed on the front page of the app, referred to as the "runway." Players all dream of receiving Top Looks and seeing their entry on the runway.

Looking at the dolls that made it to the runway as I played, it quickly become clear that while looks submitted by players vary substantially, the winning dolls usually looked very similar. Part of that relates to fashion trends in the choice of clothing, which makes sense in a fashion game. But the winning dolls also featured similar racial and ethnic looks. Specifically, unless the challenge in question had an "ethnic theme" (such as styling a Chinese bride), most of the Top Looks were fair-skinned and blond. And generally, this held true whether the inspo photo was of a pale blonde, or a brown-skinned, black-haired woman. It was a classic example of the concept of racialized "beauty queues" in fashion, where white racial features are privileged. Here's a sample of Top Looks for you to see for yourself (click to enlarge):

Voting on the looks submitted in challenges over the past year, I saw that players usually learned this lesson fairly quickly. Looks submitted by low-level players who were new to the game were much more diverse in terms of skin tone and ethnic features than those of more experienced players. New players actually styled their dolls to look like the inspo photo. More advanced players generally submitted white, blond dolls. In a "non-ethnic" challenge (for example, to style a lawyer), if the inspo model was of African or Asian descent or appeared Latina, experienced players might sometimes respond by submitting dolls with black hair instead of blond, but they still usually used skin in the accepted "white" range.

Players of all races learned their lesson. As one white player said, “I used to use the darker skin tones all the time, until I realized I was repeatedly losing events with them while I always won with the lighter skin tones.” But unsurprisingly, the beauty queues privileging white looks were most upsetting to players of color. In Facebook groups, on gaming websites, and in the fashion houses (Covet game guilds), players of color spoke about how hurtful it was to see an inspo model of color who looked like them, style a doll to resemble her, and then receive a bad score while white looks were rewarded with prizes. Players especially worried about the effect this would have on young women and girls of color who tried the game, their excitement about styling dolls that resembled themselves being replaced by shame as they learned that features like dark skin and Asian eyes were treated as "loser looks" in voting.

Even white players who thought little about race complained about the blond dominance of Top Looks. Players unlocked hairstyles that came in a wide variety of colors--browns, reds, grey, and fantasy tones of blue, green and purple. But they said they didn't feel it would be "safe" to use anything other than blond, or occasionally black or brown, unless they were styling a witch (red would be ok), a mermaid (where greens or blues might win in addition to blonds), or a space alien (in which case silver was often popular). Since most competitions were instead about women at work, on dates, or being princesses, these white players slogged through styling blond look after blond look, but griped about it.

The interesting thing to me was that most players said they wished they saw more diversity in the game. And almost all the players I encountered said that of course women of all races and ethnicities should have their beauty respected. When some African American players organized periodic events aimed at encouraging Coveters to submit dark-skinned dolls, they received plenty of likes and supportive words. And yet, despite this verbal support, players continued to vote white blond dolls the winners of competitions, and submit them themselves, saying they had to do so in order to win.

How could this cycle be broken?

The Solution?

Responding to several years of player complaints and concerns, Covet developers decided to substantially change the game in a major update. As part of a campaign they named “Evolving the Standards of Beauty,” the new version of Covet would do three things. First, it would increase the ethnic diversity of the features in the game by introducing seven new doll models with a range of skin tones and ethnic looks. Secondly, instead of employing the single fashion-doll-shaped game model based on a fantasy beauty ideal, it would make the game models come in six more realistic body shapes, from tall and twiggy to large and curvy to very petite. And thirdly, each competition would now give all players a selected base doll to style, with its body shape and skin tone locked. Players would then not have to worry that using a dark-skinned or plus-sized doll would disadvantage them in the competition, since everyone would be doing the same.

Recently Covet rolled out the update, called Covet Modern. They renamed the older version of the game Covet Classic, and allowed players to use either version for now. 

There was a lot of uproar when the update was released. In part this was because, as is common with new releases in many games, it was initially buggy, though most of those technical issues were solved within a couple of days. But many players were upset with the content of the update itself. A lot of fatphobia reared its head, but that's a matter for another post. Another vocal contingent of players claimed that they wouldn't mind styling dolls of a wide ranges of shapes and skintones, but the new ones were ugly. They said they were horribly designed, with disproportionate bodies and unfeminine features.

I should note that players in all games identify with their characters or avatars, and tend to react negatively to these being altered by the game developers. And this is particularly true of a fashion game like Covet, where one's looks are one's "weapons." Further, people are naturally attuned to facial features, and seeing "their" face suddenly change can be shocking. Rather than saying to themselves, "I am experiencing a predictable discomfort as I have just seen my game appearance change, but this will pass with time as I adjust to the update," many said, "I am upset because these new looks are objectively bad!" And they immediately switched back to Covet Classic and their familiar dolls. (The rejection of the update by many players was undoubtedly exacerbated by the very limited information they were given by the developers to prepare them for the alterations before the update was released.)

For this post about race and ethnicity in Covet, I will focus on central claim of players who objected to the update. And that is that the update "destroyed creativity" by locking skin tones and shapes for each competition, while Covet Classic had allowed players to creatively choose what to do.

Now, it's obviously silly to claim that Covet Classic did not lock the body shape of the doll, while Covet Modern does. There was only one body shape in Covet classic: the fantasy fashion-doll one. Players didn't select a shape to use in a competition in Covet Classic any more than they do in Covet Modern.

But let's examine the other part of the claim: that creativity is stifled by locking the skin color, and that this outrage against players outweighs any claimed benefit of diversity. Diversity was possible in Covet Classic--players should just be encouraged to take advantage of their options, not forced into using particular racial or ethnic features! Some said that forcing people to use all of the models was in fact the truly racist act.

How did that last claim work? There were several variations on it. One was that players should be allowed freedom to use dolls that looked like them, and for white players this meant using only white dolls, not because they were racist, but because it was simply a matter of identity (even though players are not supposed to be styling themselves, but the character in the given inspiration). Another was that by making the game model with the darkest skin have full lips and a broad nose, and the game model with the lightest skin have Asian eyes, Covet was discriminating against mixed-race people and engaging in racial stereotyping. Now, in fact, Covet Modern offers a wider range of options for styling ethnically-mixed looks, especially in the middle range of skin tones, than did Covet Classic. But some players objected that none of the darkest skin options had thin lips and a thin nose. Others protested that the facial features of the lightest-tone game model looked like a Korean pop star, doing an injustice to other Asian ethnicities (though there are Asian eye styles available in the make-up sets for the other skin tones that have a variety of eye shapes). Yet others said that requiring anyone to use the palest skin tone was racist, because it was overly white, privileging fairness. In fact, Covet Modern changed the skin tones, removing one of the lightest options and offering more skin colors in the middle range, so that the options offered shifted away from very light skin. But it is true that the lightest skin tone in Modern is still very pale--one could argue, so pale as to inspire idealizing of extreme fairness (or goth makeup).

What I do have to note here is this: a whole lot of the players making arguments that the new Covet models were offensive to people of color were white people. Were these people stepping in as white allies supporting arguments of people of color, or positioning themselves as antiracist to excuse their stance against increasing diversity? In the wider world, it is now common for people who are opposing antiracist movements to frame themselves as the true champions of racial equality. But perhaps these claims were true, and Covet Modern was having the unintended consequence of decreasing the diversity of looks and further encouraging the preference for pale, blond looks or racial stereotypes generally.

What would help here is some evidence about whether the update is having the desired effect of increasing the diversity of looks submitted, and what the impact seems to be on creativity. So I did a little analysis, which I'll share here. I examined the top looks in four contests that took place within the first week of the update release, and analyzed the top looks in both Covet Classic and Covet Modern.

The Results

Let's look first at the competition Sunday Brunch, in which players were instructed to style a college student meeting her grandparents for a meal at their country club. The inspo model appears to be Latina or of mixed ethnicity, with long dark hair, dark eyes, and skin in the Covet medium-dark range. And here are representative Top Looks in Covet Classic and Covet Modern that received many likes (I've chosen two with the same outfit to hold that factor constant for you):

At first glance it is obvious that the two Top Looks pictured are very different. For one thing, the Classic doll is very tall and thin, while the Modern model assigned for this challenge is shorter and curvier. But for purposes of this post, let's just look at their disparate ethnic features.

I examined each of the 62 Sunday Brunch Top Looks in Covet Classic and the 60 Top Looks in Covet Modern, and noted the skin tone and hair color used in each of them. First, some information about skin tones. In Covet Classic, there are 6 skin tones, referred to as T1 through T6. The two palest, T1 and T2, are rarely found in Top Looks unless the player should be styling a ghost. Tones T3 and T4, the midrange, are used in most winning looks. T3 is used for a fairer white look, and T4 for tanned white looks, or sometimes to indicate some degree of ethnic ambiguity. T5 is found occasionally in Top Looks. It is the skin tone most frequently used when styling the doll in an "ethnic" challenge calling for dark skin, when the challenge is "exotic" in nature, or for very tan white looks. T6, the darkest Classic tone, is found rarely in Top Looks. When a T6 classic doll does make it to the runway, it is usually in a challenge to style a historical character of African descent--but it is also problematically popular when styling monster characters and evil practitioners of magic, even if the inspo photo pictures a white model.

Where Covet Classic has these six skin tones, Covet modern gives players seven models, each with a different skin tone. The T1/T2 pale range of Classic has been replaced by a single pale model with Asian eyes. The second palest tone, with European features, corresponds to Classic T3. The seventh, darkest model, with African features, corresponds in tone to Classic T6. And there are three skin tones between these in Modern where Classic has two, giving more options in the medium-dark range.

OK then. Those are the skin tones in the two versions of Covet.

In the Covet Modern version of the Sunday Brunch competition, the skin was locked on the second darkest Modern tone, so 100% of the entries had fairly dark skin. What skin tones were found in the Top Looks in Covet Classic? The answer: one Top Look used the very pale T2, and the other 61 used the "standard white" midrange of T3-T4. There were no Top Looks in Classic using a skin tone as dark as the locked skin in Modern.

Let's also look at the hair colors used in the winning looks. In Covet Classic, 81% of the Top Look dolls had blond hair. Brown hair was used in 18% of the Classic Top Looks, black hair in none of them, and 1 doll used light red hair. Compare this to the hair colors found in the Covet Modern Top Looks. Blond hair was used only 5% of the time. Brown hair appeared 38% of the time, and 57% of the Top Looks had black hair. The hair found in the Modern Top Looks was much, much darker than in the Classic Top Looks.

So, over 80% of the time, the Top Looks had white skin and blond hair in Classic. This is the white-privileging, usual outcome in Covet Classic challenges, found, as we see here, despite the fact that the inspo photo featured a woman of color. On the other hand, all of the Top Looks in the Modern update had darker skin, and only 3 of the 60 were blond--each of these featuring not pale blond hair, but bleached ombre styles with dark roots and light ends. Thus, we see a very stark difference in the ethnic outcomes of the Sunday Brunch challenge. The Classic winning looks appeared very white, while the Classic top scores were equally emphatically not white looking.

Furthermore, there was more diversity of hair color used in the Modern Top Looks, with the favored hair color (black) accounting for only 57% of all looks, in comparison with the 81% of the Classic Top Looks using the hair color favored there (blond). More diversity of looks implies more creativity in styling.

But perhaps this outcome was a fluke. Let's look at another, similar challenge.

The competition entitled "Five-Star Views" asked players to style a wealthy woman staying in a 5-star Paris Hotel, about to take a chauffeured car to dinner with her husband. The inspo model again had a Latina or mixed-race look. Her skin appeared to be somewhat darker than that of the inspo model in Sunday Brunch, and her wavy hair highlighted to appear a bit lighter. Here are representative Top Looks in the two Covet versions:

You can see that the outcomes give an impression very similar to that in Sunday Brunch. This is interesting, because in Five-Star Views, the skin tone assigned in the Modern competition was not as dark as that set for Sunday Brunch. It uses the skin color in the middle of the Modern tone range, which is a bit darker than the Classic T4. And almost half of the Classic Top Looks, including the one pictured above, used the Classic T4 "olive" tone. (One Classic Top Look used the very light T2, one used the dark T5, and half used T3, the "standard untanned white person" tone seen in the majority of Sunday Brunch Classic Top Looks.)

So, the average Classic winning look was slightly darker than in Sunday Brunch, while the Modern skin tone was lighter, meaning that the skin tones were more convergent in Five-Star Views. This was mostly due to the tone locked in Modern being lighter, but the greater use of T4 in Classic contributed. And yet the the winning dolls look clearly differentiated, as the photo shows.

Obviously, a major reason is the use of blond hair in Classic Top Looks. 87% of the Classic Top Looks used blond hair, which is even higher than the 81% in Sunday Brunch. (Brown hair and black hair were each used 7% of the time.) Now, Modern Top Looks in Five-Star Views were also more likely to feature blond hair than they did in Sunday Brunch--it was used 21% of the time instead of only 5%. The fact that all of the winning looks employed gold formal dresses, echoing the dress worn by the inspo model, most likely explains the increase in use of blond hair in both Classic and Modern, pairing golden locks with a golden gown. Still, most Modern Top Looks employed darker hair--52% brown and 24% black. And that has a huge impact on the ethnic appearance of the outcomes.

But another factor plays in as well, and that is the facial features of the dolls. These are selected by the players in choosing their make up options (the equivalent to character "skins" in other games). So we need to talk about Covet make up. In Modern, each of the seven skin tones has its own unique model, with specific ethnic features. The makeup options for each model are unique to that skin tone, designed with the intention of looking good on her skin tone and features. The ethnic appearance of the doll can be shifted somewhat depending on the make up chosen, especially for the mid-range skin tones. The model assigned for Five-Star Views, for example, can be made to look more Asian, more Latina, more Native American, more European, or more African in ancestry depending on the selected make up. Her appearance is particularly ethnically flexible and often gives a general impression of mixed ancestry, with her nose being wider than that of the Classic doll, and her fairly generous lips. In the Modern Top Look I've included, the makeup used makes the entry look a lot more like the inspo model than the Classic comparator, not just because of the red lipstick, but because of the ethnic features of the doll's face the make up is designed around.

The Classic doll, on the other hand, does not have ethnic variations for differing skin tones. Her ethnic appearance varies only due to the make up chosen by the player, so there is a narrower range of ethnic variance. For example, players can unlock a total of 14 Classic make up options with Asian eyes as they level up. In contrast, the Modern lightest model has Asian eyes, and a current total of 58 make up options across all of the levels unique to the Asian model. There are also Modern make up options for most skin tones that can look Asian. In particular, the second-darkest model has many South-Asian-appearing make up options, including several featuring bindi ornamentation. So the Classic Asian options are much more constrained than those in Modern. Furthermore, it's not just the smaller number of "ethnic" make up options that limits Classic racial and ethnic styling. Since the make up in Classic is the same no matter which skin tone it is applied over, many of the makeup options that are not specifically ethnically marked only look good on a few skin tones, usually the midrange "white" ones. Classic players often complain that few make up options look good on the darker skin tones, and in fact, cite this as a reason that they stick to the midrange T3/T4 skin tones.

Interestingly, these constraints on ethnic styling options tend to make Classic players view subtle variations in midrange-skintone looks through an ethnic lens. If you are unfamiliar with Covet you might not know it, but a substantial number of Classic Top Looks, including the one in the illustration above, in fact employ make up options which are read by many players as appropriate for "ethnic" looks, as well as youthful "edgy" white looks. This subset of Classic make ups all have open, pouting mouths will medium-full lips, and evoke the look of Kim Kardashian after a session of lip-filler work. (These make up options are considered highly desirable in Covet Classic, and since most of them are unlocked only at quite high levels, an incentive to level and a significant element in high-level player advantage.)

In fact, if you look at the pictured Classic Top Look in Five-Star Views using the eyes of an experienced Covet Classic player, the player's choice of tan T4 skin, a plumped "exotic" mouth, and thickly waving hair (albeit blond) reads as giving a nod to, or even embodying, an "ethnic" rather than white look. Taking off that lens though, and looking at the Classic doll objectively in comparison to the Modern Top Look, it's a very whitewashed vision of a nonwhite look.

Consider the possibility that the player who submitted this Classic doll wanted to create a style that resembled the model ethnically, and chose a make up option with one of the pouty mouths viewed as ethnically ambiguous for this reason. The blond hair choice could have been a reluctant one, believed necessary in order to have a shot at a Top Look, and the choice of the heavily waved hair an attempt to offset the blondness with another note of ethnic ambiguity. If this was the player's intent, the resulting look is sad, because while it has scored a Top Look, it doesn't actually look much like the inspo model. In that case, the Modern update is doing its job. Covet Modern allows players to create looks with facial features that are more ethnically representative, and this seems to free them to also feel secure using a wider range of hair colors, while simultaneously skewing much less toward blondness. The white, blond imperative is much reduced, and ethnic diversity enabled.

Let's move on, then, to look at a different challenge--one featuring an inspo model who is dark-skinned and clearly of African ancestry.

For the Training Day challenge, players were asked to style a boxer training for a charity match. The inspiration model is dark-skinned, well-muscled, and has hair that appears to be narrow cornrows or braided extensions pulled back in a ponytail. Here are the representative Classic and Modern Top Looks (again paired for similar outfits):

I've stated that Classic players rarely used the dark range of T5/T6 skin tones--yet in this competition we see that the mostly commonly selected skin tone in the Classic Top Looks for Training Day was T5. The breakdown: 24% used T3, 16% T4, and 60% T5 (no Top Look used T1, T2, or T6). In terms of Covet Classic competition outcomes, the average Training Day Top Look was unusually dark. Still, no Classic Top Look used T6, while the Modern players were given the darkest model to style, so the Modern outcomes all looked more like the inspo model than the Classic Top Looks did.

Still, why would Classic players use a darker skin tone in this competition, when they ignored the skin tone of the inspiration model in the challenges we looked at earlier? You might suppose that it is because the Training Day inspo model simply has substantially darker skin--but I have seen other competitions in the past featuring very dark-skinned models where the Top Looks were all fair-skinned. I believe it is the narrative description rather than the model's appearance that is determinative. You may recall that I stated that Classic players often use T5 for competitions about warriors, monsters, evil shamans--and brown-skinned athletes, which is what players were asked to style in this case. Stereotypes relating dark skin to physical prowess and dangerousness are in operation.

I also believe that class plays a substantial role here. The Sunday Brunch and Five-Star Views competitions were about wealthy individuals (as you don't find many poor people eating at country clubs or staying in 5-star Parisian hotels). Wealth and "classiness" are associated with whiteness in a positive stereotype, which seems to deter Covet Classic players from submitting dark-skinned, dark-haired looks when styling a wealthy character, even when looking at an inspo photo showing a woman of color in that role.

The fact that the boxing match in the competition narrative is supposed to be for charity may also play a role. Evoking charity, social justice, and activism can put players in a mindset that makes them more willing to use a darker skin tone when faced with a black inspiration model. This might have been especially true at the time of the Training Day competition, because the Black Beauty Blackout was in progress. This is an annual event in which players who advocate for racial equality urge other players to submit looks employing Classic tones T5 and T6 in all their entries for the week. Mind you, that event was taking place during the other competitions we've examined, with no apparent impact whatsoever on the resulting Top Looks. Of the 121 Classic Top Looks in Sunday Brunch and Five-Star View, exactly one employed T5 skin and none used T6, despite the inspo models being women of color and the Black Beauty Blackout being in progress. But perhaps the Blackout event had more impact in a situation where the inspo model was of African descent and the scenario involved punching people for charity.

Looking at the Training Day comparison Top Looks, it's clear that while darker skin than usual was used, Classic players still did not vote looks using the darkest tone to Top Look status. In Modern, all the top looks used the darkest model, whose features include a wider nose and lips in a range of classically African ethnic looks, making the Top Looks appear "blacker," not just in color, but in ethnic appearance. But there's more to the difference than that--again, hair plays an important role.

Just as the average skin tone used by Classic players in Training Day was darker than usual, so were the hair color choices. Classic Top Looks used blond hair "only" 33% of the time, brown hair 16% of the time, black hair 49% of the time, and one player chose white hair. While this is a darker average than typical in Classic, it is still skews substantially lighter than the haircolor choices in the Modern Top Looks. In the Modern Top Looks we see dark blond hair 5% of the time, brown hair 26% of the time, and black hair 66% of the time, with one player using auburn red hair.

The general impression one gets looking at the Training Day Classic Top Looks is of racial and ethnic ambiguity. The Modern winners look like black women. The Classic ones. . . well, considered in the context of Covet Classic traditions, they are dark. But they deploy--I believe deliberately--racial and ethnic signifiers in a mixed manner. In the example I've shown above, we see T5 medium brown skin paired with hair the game names dark blond. Classic Top Looks looks featuring black hair were more likely to be paired with fair skin. The ethnically ambiguous Classic make up styles with Kardashianesque lips were everywhere.

My read of this is that Classic players are are often trying to straddle two goals in competitions like Training Day. The first is to style a look that could be read as black (either because they desire this personally, or because they fear a white look would be downvoted by offended players of color, given the race of the inspiration model). In tension with this goal, players also wish to harness the privileging of whiteness in the game. I've felt the pressure to do this myself as a player: not to make a Classic doll look "too black" by using the make up with the broadest lips or the dreadlocked hair options, lest I get a poor score and disappoint my fellow fashion house members who are relying on each of us to do well for "rallies" (the Covet equivalent of guild raids). I know that I had to actively work to stop myself from unconsciously using the Asian makeup and darkest T6 skin less and less. The pressure is there to style "ethnic" models as Rorschach dolls, onto which any voter can project what they wish to see, in terms of race.

These player concoctions of ethnic ambiguity were very effective, by the way, and the evidence for this is that I found it unusually difficult to determine by looking exactly which skin tone a Classic Top Look in Training Day had employed. I often wound up having to screenshot them, crop samples of exposed skin, and export them to a color swatch app to compare to reference tones to be sure which skin tone had been used.

And I believe this relates to the claim some players make when opposing Covet Modern--that it "limits creativity." It's true that styling an ethnically ambiguous doll takes experience and finesse. But this is personal effort being unfortunately expended in order to avoid producing clearly racialized dolls of color. Of course, women of every conceivable racial admixture exist, and deserve representation. But a whole lot of real people look clearly Asian or African in their features, and dolls with those features are rarely seen on the runway in Covet Classic, while ambiguous dolls are seen much more frequently. By presenting players in Covet Modern with 7 models bearing distinctive ethnic features, this lack of representation in winning looks is addressed. (Yes, most Modern models can have their features shifted in ethnic looks via make up choices, but the lightest tone mostly looks Asian, the second-lightest mostly European, and the darkest mostly African.)

I've also heard some players saying that they like to style looks that appear mixed-race, and they are angry at Covet Modern because it "enforces racial stereotypes" by making it harder to produce mixed-race looks. But looking at player submissions in Modern across competitions, one sees so much more ethnic diversity, including many dolls that can appear to be of mixed race, that this argument seems strange. It may be that many Classic players have deeply if subconsciously internalized the belief that being able to style ethnically ambiguous dolls is the key to winning competitions when a white blond look might be penalized by offended players, and that this produces their fear of sometimes being forced to use more strongly racialized models in Modern.

Personally, I find it a great relief that Covet Modern has removed from the game the pressure to create ethnically ambiguous looks in order to win prizes while styling dolls of color. Diversity is locked in, and styling dolls using every skin tone and ethnic appearance is a pleasure untainted by fear of being penalized by voters.

So, we have looked at what a difference is made by the Covet Modern update in competitions featuring an inspiration model of color. But another question remains: are competition results affected when the inspiration model is white?

Let us consider the Petals and Jewels competition, in which players were asked to style a princess celebrating her birthday in a garden setting. The inspo model was a fair-skinned woman of European ancestry in a classic Western Renaissance pose. She had brown hair. Here are representative Top Looks:

Examining the Classic Top Looks, we find no surprises: 98% used the "white standard" range of T3/T4, while one doll used the pale T2. And 89% had blond hair, 10% black hair, and a single doll had brown hair. There was white, blond hegemony (despite the fact that the inspo model was brunette).

The results in Covet Modern were more interesting. All of the dolls were based on the third lightest model, whose skin is close in tone to Classic T4, so the Top Looks were about the same shade in both versions of the game. But the compulsion to use blond hair was substantially less. 52% of the Modern Top Looks had blond hair, 25% brown hair, 21% black hair, and 1 was a redhead.

So, even though all of the looks were in the light medium range, there was a lot more diversity in the Modern Top Looks. "Only" half of them were blond. 

In Covet Classic, there was technically total freedom of skin color and thus more potential diversity in a given single competition. In fact, because I was participating in the Black Beauty Blackout, I was switching between versions, styling in Modern unless a light skin was locked in, in which case I'd return to Classic to submit a T6 look. Here is my submission:

No dark-skinned princess look styled with a skin tone like my submission made her way to the Petals and Jewels Top Look runway. So, in a way it is true that there is more freedom of styling in a given competition in Classic, in that all players are free to select any skin tone. But in practice, to break the white, blond rule is usually to lose, and pragmatically speaking, using skin tones T3 or T4 in styling a doll based on a white inspo photo was treated as mandatory by most players.

The result was that Classic Top Looks locked in light skin and came close to doing the same for blond hair. In Modern, however, only the skin was locked, and players used a wider variety of haircolors.

Why might the Modern players be submitting more diverse looks? One possibility may be that, at this point in time, the Covet Modern player base is distinctive. A majority of players have not switched over, and those who have may be more open to changing how they play the game, and/or more motivated by the goal of increasing diversity in looks. Another possibility, however, is that players who have gotten used to using Covet Modern have been influenced by submitting a much wider variety of model looks, so that the white-and-blond-is-best rule is being purged from their styling and voting routines. The inspiration model not being blond, half of the looks winning their place on the runway were not blond either. If true, this means Covet Modern is working as advertised. It is opening players up to the full range of beautiful looks--including greater diversity in white looks.

Covet Modern Works

We have seen that Covet Modern was introduced in order to enact what players themselves said they wanted: more diversity in looks. Players asked for plus sized models and for more ethnic diversity and representation in the game. Players had stated that dominion of white, blond looks was boring. 

What could be done to change things? One tactic was to try to convince players to change their styling and voting behavior to reflect their own stated beliefs that all women have the right to be found beautiful. An example of a player-side effort in this vein was the Black Beauty Blackout. As we have seen, although such efforts received praise from players, they achieved little in terms of outcomes. 

Why didn't these voluntary educational and advocacy efforts have the desired effect? One reason is that there can be a disjunction between what many players say out loud, and what they personally believe, but don't say. Players may be actively, consciously racist, but enact this only in their styling and voting behavior, not in what they say to others. (There are of course some Covet players who willing choose to say things they know to be overtly racist about beauty and ethnicity in public forums, but they are moderately few.) 

While this conscious, if quietly held, bias surely explains some Covet player behavior, it is equally surely not the whole story. Another, especially likely possibility is that player biases are largely unconscious. Perhaps you have heard of "implicit bias" experiments. One example of these involves having research subjects view a series of photos of black and white faces paired with positive and negative words (such as "nice" and "mean"). Sometimes they are asked to press a button when a black face is paired with a positive word, and other times when a black face is paired with a negative one. People of all races push the button more quickly when assigned to press it when they see a negative word paired with a black face than they do when told to press the button when a black face is paired with a positive word. Conversely, people of all races press the button more quickly when asked to pair white faces with positive words than they do when told to pair white faces with negative words. We are socialized to associate whiteness with goodness and beauty, and blackness with badness and ugliness. People of all races unconsciously internalize this message, even when they are consciously committed to antiracism. And this influences their actions, such as pushing a button.

Covet voting asks players to basically push a button selecting the "better look" of two dolls, which players do over and over again. This makes it very similar to an implicit bias study. Few voters spend a long time studying the two options presented and making their selection. They glance and vote, glance and vote, and probably engage in little introspection about what besides outfit choice may be influencing their decisions to tap the look on the right or the look on the left.

We can imagine that in Covet Classic, the implicit bias factor would lead to a mild preference for whiter looks in the early days of the game. As players viewed at the Top Looks, and saw more blond, light-skinned dolls there, this preference would be reinforced subconsiously. This would lead players to vote for whiter looks, and seeing more of these on the runway would further increase the unconscious bias, producing a feedback loop leading to an ever greater advantage for light-skinned, blond dolls. 

Bear in mind that unconscious biases are often independent from conscious beliefs. Thus, many players would consciously say that they valued diversity and wanted to see plenty of it, while subconscious factors led them to move in the opposite direction, behavior-wise. This makes attempting to influence their behavior through conscious, verbal appeals to valuing diversity fairly useless. Players can nod and agree, while their behavior is unaffected.

We like to believe that our behavior is directed by our conscious beliefs, but quite often, the exact opposite is true. And one result of this is the negative effect of moving through a world or game in which racial inequality is operating: by following the rules in our behavior, we come to internalize them in our hearts. We know that by the time they are four, most American children of all races prefer to play with white baby dolls over black ones, and identify the white doll as the pretty one and the black doll as the "bad" one. While this is hurtful to everyone on an ethical or spiritual level, it is particularly damaging to children of color.

Covet Classic had come to function in such a way that it made this social problem worse. What I posit began as an unconscious, implicit bias became something consciously understood as a "rule of the game." A conspiracy theory grew and flourished that voting did not actually determine outcomes, but that dolls were being evaluated by some Covet "algorithm," and that algorithm included the "right" skin tone for a Top Look, which was usually T3 or T4. Resistance was futile. It was framed as a silly, idealistic, unrealistic thing not to face the reality that white, blond looks win competitions. Anyway, players said, it was just a game--not a big deal. If you don't like it, just don't play. Or keep having your idealistic events asking players to use dark skin, and if people want to, they can make their little gesture, but nothing is really going to change. Even if change would be nice--which of course, everyone agrees it would be--said players.

Something really needed to be done to change this state of affairs--something that actually worked, where appeals to conscious ethical action had not.

So Covet Modern was developed and released to address the problem, and it is indeed working. Locking the skin color for each competition quickly produced a great diversity of ethnic looks. And while it is early days yet and the update is young, there are signs that it is increasing the diversity of player choices in styling white looks, not just when styling dolls of color. Creativity seems to be supplanting a boring rut of seeing runway after runway filled with light-skinned, blond-haired dolls wearing the same small subset of make up options.

This is by no means to say the update has been perfected. I see problems with the new user interface. The question of whether the lightest skin tone is unrealistically pale is a valid one. The facial features of the second-lightest, European model were quickly redesigned in response to an outcry from players objecting to her ethnic look of thinner lips, pale eyes, and a sharper nose, which many claimed made the white model look ugly, so that she now has unrealistically plump lips and smouldering eyes, resembling the favored looks in the Classic game. While many players thought that was a big improvement, it reintroduced issues of unrealistic bodily ideals--in this case a plumpness of lip few white women embody without plastic surgical intervention. And there is a very big problem with game clothing not fitting the new curvy models well. Items with patterns and ornaments were imported into the game by Photoshopping them to fit the Classic doll, so the patterns and details look distorted when the items are used to style a plus-size doll.

But Modern is, in my eyes as an experienced, top-level player, a big improvement. It is causing me to be more creative in my styling. The game feels much more fresh, much more challenging--and much less problematic. Playing Covet Classic for a year has felt part lighthearted fun, and part ethnographic foray into a disturbing realm that was inculcating racial bias and damaging, unrealistic bodily ideals. Playing Covet Modern, on the other hand, feels like it is building racial and size empathy as one tries to make each model look her best. It feels much more healthy to me. (I should specify: not in the sense that replacing a scrambled egg breakfast you love with oatmeal you despise is healthy. It's much more like finally getting a delicious fruit salad after months of being given bland marshmallows as one's daily "treat.") And, fears of high-level Classic players notwithstanding, my scores have not suffered as a result of switching to Modern.

I hope that the many players who rejected the update at first glance, or never tried it because they heard it was "bad," can be convinced to immerse themselves in playing it for a while. There is an adjustment period--as there is to any change. But if players give Covet Modern a fair chance, many of them may come to realize that it is personally rewarding as well as more supportive of the self-esteem of all players.