A Cosplay Photo Set in the Esports Feed: When a Misclassification Becomes a Traffic Machine

6 a.m. Paris time. I open my esports feed, coffee still hot on the desk facing the balcony, and the first thing that appears is not a scoreline, not a balance patch, not a transfer. It is a cosplay photo set. A girl in a stylized sailor outfit, artificial rabbit ears pinned to her hair, standing in an elaborately staged setting, recreating a character from Azur Lane — a mobile gacha game where players collect girls who are the embodiment of warships. The photo set is tagged "esports." And right below it, in the related-links block, is an entirely different headline: a Vietnamese PUBG player facing a possible ban. Two worlds. Two systems of logic. The same frame, the same label. I have spent fourteen years covering this industry, from the days of grinding ranked in Paris to the nights spent dissecting every pick-and-ban. The match begins when the coaching staff submits the roster, not when the referee blows the whistle — that is the line I still tell my journalism students. And the moment a cosplay photo set landed in an esports feed made me stop. Not because the photos were bad. They were carefully invested. But because their presence there, under that label, says something about how we are redefining this industry — and about the numbers we use to measure it. Before going further, a classification step is required. Azur Lane is not an esports title. It is a character-collection gacha game, operated by Manjuu and Yongshi, and it has no meaningful professional tournament system. There is no franchise league like the LEC or LCS. There are no regional qualifiers. No group stage. No world champion. Its content cycle is decided by character banners and new outfits, not by balance patches. In the world of Azur Lane, what decides revenue is not win rate, but the recognizability of a character. The character in this photo set is Shimakaze, a Sakura Empire destroyer in the game. That name matters. In the fan community, Shimakaze is one of the most recognizable faces — a clear design, easy to recreate, and above all, the ability to transform through countless outfits. That is a marketing trait, not a balance trait. A character who can wear ten different outfits is a character who sells ten times. The photo set was made by a cosplayer, and the article describes it with familiar adjectives: carefully invested, impressively transformed, close to the in-game version, a mischievous spirit. There is not a single number. No views. No engagement rate. No independent measurement. Only carefully placed praise — the familiar mark of a promotional placement, what editors in Vietnam jokingly call a "placed piece." No one is wrong. No one has violated anything. There is only a piece of content born to please, filed into a drawer that does not belong to it. So why does a cosplay photo set sit in an esports feed? The answer lies at the intersection between the economics of a gacha IP and the economics of media traffic. Start from the publisher's side. A gacha game lives on a very specific content flywheel: character design creates fans, fans create user content, user content nourishes brand recognition, and brand recognition sells outfits back. Cosplay sits exactly at the second link. When a cosplayer recreates Shimakaze, they are not merely making a piece of art — they are making a piece of content capable of spreading across language borders. A good photo set can travel from a Japanese-speaking fan community to a Chinese-speaking one, to the West, and to Southeast Asia, all through images alone, no translation needed. No language barrier stops a beautiful photo set. This is what I call the gacha-IP flywheel. It is a marketing machine parallel to esports, not a branch of esports. And it runs on a completely different calendar: the calendar of banners and outfits, not the calendar of a season. A tournament has an opening day and a closing day. A character banner has only a window that everyone knows will end, and when it ends, the flow of content moves on to another character. In that world, there is no such thing as training or form. Only which character is getting attention. There is one more legal layer we usually overlook. Cosplay exists in the tolerance zone of copyright. Gacha game publishers in general, and Azur Lane in particular, tend to be lenient toward fan works — because they understand that every cosplay is a free advertisement for their IP. But that leniency is not a right. It is a permission that can be revoked at any time. And in the same Vietnamese gaming-media market, copyright disputes between parties still simmer quietly, reminding us that the line between fan work and infringement is a faint one, redrawn by the rights-holder at will. Now look from the newsroom side. A gaming site in Vietnam earns money through traffic and display advertising. Real esports news — transfers, punishments, coaching changes, contract disputes — requires sources, relationships, and time to verify. A cosplay photo set does not. It is ready-made, it is visually appealing, it cannot be factually wrong because it has no facts to get wrong. It only needs a catchy headline and a correctly placed classification tag. Economically, this is a perfect deal: near-zero cost, near-zero risk, and far higher engagement potential than a dry transfer line. And here is the crux. That "esports" tag is not a deliberate editorial choice in most cases. It is a by-product of the system. When a site publishes both PUBG news and cosplay photos, the keyword-based and content-neighbor tagging algorithm automatically files both into the same topic cluster. The keywords "game," "character," "community," "trailer" appear in both. So the cosplay photo set gets pushed into the esports stream. No one pressed a button saying "this is esports." The machine did it, and the machine cannot tell a match from a photo set, because both have color, images, and names beloved by a community. The effect goes deeper than a wrong tag. In my years working with industry data, I learned one thing: dirty data is not in the wrong numbers — it is in the things that are right but filed in the wrong place. A cosplay promo tagged as esports inflates the esports content volume of a market. It skews every subsequent metric: topic trends, interest levels, industry growth rates. If an analytics platform in Vietnam reports that esports content grew forty percent in a quarter, I want to know how much of that is cosplay photos. Because if the answer is a third, then that forty percent growth is not the growth of esports. It is the growth of another industry wearing a name. We are used to the line: people call it meta, I call it digitized fear. But here there is another version, and it is more uncomfortable: people call it esports, I call it a traffic container. And that container is swelling with content that has no opponents, no tournaments, no winners and losers. There is no match at all. Only images of characters someone spent weeks stitching thread by thread. To see more clearly, let us put the two pieces of content sitting side by side in that morning's feed on the operating table. The first is the Shimakaze photo set. Its value lies in aesthetics and fidelity to the source. The reader consumes it with a fan's eye: does the character look right, is the outfit detailed, is the expression faithful. This is an exchange within a community — a broad community, but a community nonetheless. There is nothing wrong with this kind of content. It is part of fan culture, and fan culture deserves respect. The second is the news about a Vietnamese PUBG player facing a possible ban. Its value lies in consequence: a publisher's decision, a dispute between parties, a sanction that can change a person's career and the landscape of a region. The reader consumes it with a competitor's eye: who is right, who is wrong, is this decision fair, how does it affect the upcoming tournament. These two are not the same species. One is fan culture, one is competitive governance. One exists because of character recognition, one exists because of rules and results. One is born to be admired, one is born to be argued. And if you force these two into the same label, you do not enrich the label. You dilute it. You turn a definable category into a box into which anything can fall, so long as it relates to a video game. I remember once in Paris, at a press conference after a final, a French colleague asked me whether esports is really a sport. I answered with a question in return: is a meal really a meal if people call everything edible food? The answer lies not in definition, but in classification. And we are classifying very badly. But wait. I must be careful here, because the instinct of an intellectual provocateur is to reverse every conclusion, and in this case, the reversed conclusion is more correct than the original one. All of us — esports journalists, analysts, those who consider ourselves keepers of standards — react the same way when we see a cosplay photo set in an esports feed. We grimace. We talk about data contamination. We call for re-tagging. We treat it as a mistake to be fixed, a carelessness, a thing to be cleaned out of our tidy house. But if we look more closely, we see that tagging that photo set as "esports" is not an accidental error. It is an accurate reflection of a truth the esports elite does not want to admit: that "esports" was never a clean category. It is a marketing umbrella stretched over years to gather very different things in one place. Think about how that umbrella was built. When major tournaments need sponsors, they do not sell a tournament of one game. They sell an audience that is young, tech-savvy, and has spending power. The label "esports" was born to sell that. It does not describe a discipline. It describes a demographic. And when you sell a demographic, you do not care whether people come to your content for a brilliant play or for a beautiful photo set. You only care whether they come. In that context, the Shimakaze photo set is no longer an intruder. It is a legitimate member of the "esports" house — if we admit that house was never built on a strict definition. It was built on traffic. And this photo set carries traffic. I do not say this to excuse carelessness. I say it to be precise. Because there is a difference between a label being misunderstood and a label never having meant what we thought. The cosplay case belongs to the second kind. And the irony is that we, the ones complaining about it, are the ones who helped build that umbrella. And if I am wrong? If "esports" truly has a strict definition, and this photo set really is an impostor? Then the next question is just as uncomfortable: who has the right to define it? A committee? A publisher? A newsroom that needs traffic, or a journalist who needs to keep professional face? Ranking is only the way people recount what they have not understood — and here, the ranking in dispute is a definition. What is notable is that in this particular case, no one is harmed. The cosplayer gets attention. The newsroom gets traffic. Azur Lane fans get one more piece to admire. The PUBG community has its real news, just pushed near the bottom of the page. The only harm, and it is an invisible one, is to the quality of the analytics industry — to people, like me, who try to read the signals of this market and need their data clean. An investor reading a report on Southeast Asian esports growth will not know that part of that number comes from photo sets. If we buy a single thread and call it the whole cloth, we will sew a shirt that fits no one. So what should we do with misplaced cosplay photo sets? I do not think the answer is to kick them out. I think the answer is to label them correctly — and to accept that in doing so, we are admitting "esports" is a bigger umbrella than we take pride in. Because the stadium is empty, but I still hear the crowd that never came — and that crowd does not come only for teamfights. They come for a community, an aesthetic, a sense of belonging. That is a fact, not a mistake. If this industry wants to be taken seriously, it must do something it keeps avoiding: separate the layers. One layer for competition. One layer for fan culture. One layer for marketing. Not to exclude each other, but to understand each other. A cosplay photo set does not cheapen a match. But a statistic that conflates the two cheapens both. And if we cannot do that, then ten years from now, when someone asks what shaped esports in Southeast Asia, the answer may not be patches or finals. It may be photo sets — and the numbers that counted them as if they were matches.