Shimakaze Next to a PUBG Headline: The Mislabeling Eroding Vietnam's Esports Feed

I opened my esports feed on a Tuesday morning, and what sat between two headlines about PUBG Asia Stars was a photo set. A young woman in a red-and-white sailor outfit, rabbit ears standing up, posed on a mock ship deck under green backdrops and soft light. The caption said a Chinese cosplayer, known by the handle Tie Shou Jiao Shou, had just released a set embodying Shimakaze, a character from Azur Lane. The photo set sat in the same section as news that a PUBG player might be suspended.

No tournament. No scoreboard. No contract. Nobody competing.

I sat still for about ten seconds, then did what I always do: opened a blank spreadsheet and started counting. Of the forty most recent headlines in that section, nineteen were about a match, a contract, a sanction, or a standings table. The other twenty-one were about characters, skins, cosplay, a photo set, or a short clip with nobody competing. That number made me pour another coffee. But the question that bothered me more than the number was a different one: if I call this an error, what am I protecting, and whom am I protecting it for?

Azur Lane launched in 2026, developed by Manjuu and Yongshi. It is a gacha character-collection game. Players do not climb a ranked ladder in the sense of League of Legends or CS2. There is no licensed professional tournament circuit, no regional qualifiers, no franchise slots, no teams signing long-term player contracts, no seasonal salaries.

What runs Azur Lane is a banner schedule. Every few weeks the publisher opens a new character banner or reruns an old one, usually bundled with an outfit. Revenue comes from players spending money for a probability of getting the character or skin they want. Its content cycle is not measured in balance patches, but in banner rotations and outfit releases. When a balance patch lands in a fighting game, it changes how players optimize. When a banner lands in Azur Lane, it changes how fans spend and how they talk to each other online.

Shimakaze exists inside that logic. She is a destroyer of the Sakura Empire in the game, and she is famous for a design that is easy to recognize — rabbit ears, a sailor outfit, a blend of warrior imagery and cuteness. That is a design built to spread. The rabbit ears make the character identifiable in a small thumbnail on a phone. The sailor outfit lets her stand in many different photo settings. The warrior-cute blend keeps her from being boxed into a single audience. In the fan-content economy, a recognizable character is an asset that can pay out repeatedly.

And here is the setting of the feed I am talking about. A Vietnamese-language site aggregating multiple titles, running character content and competition news side by side. The author of the photo-set piece signed as Tuan Hung. Right beside it were headlines about PUBG Asia Stars, about a Vietnamese player named Himass facing a possible suspension, about an apology connected to KRAFTON, and about a dispute between two figures in the community. Two different worlds sitting side by side in one column.

In seven years of watching esports feeds from Seoul to Los Angeles, I have grown used to sections being mixed. But there is a difference between mixing out of curiosity and mixing out of algorithm. A newsroom mixes content because it believes its audience cares about both things. An algorithm mixes content because both things share a keyword. The result on the screen looks the same. The cause is entirely different.

How a cosplay piece ends up in an esports section is no mystery. It passes through four doors.

The first door is the keyword. The headline contains the name of a game. The site's automated classifier, taught that a game name plus the word game belongs to the game category, folds the game category into the esports category as two overlapping sets. A piece about Azur Lane gets pushed into esports because Azur Lane has the word game inside its definition.

The second door is the recommendation link. Cosplay pieces get placed next to esports pieces because the system sees that readers of one read the other. That correlation is true behaviorally, but false topically. Azur Lane players and PUBG viewers may be the same person; that does not turn a Shimakaze photo set into competition news.

The third door is traffic economics. A high-quality cosplay set has a high click rate and long time-on-page. For a site living on display advertising, that is valuable inventory. A story about a player suspension has a lower click rate and a shorter lifespan. If both sit in one section, that section looks livelier on the dashboard.

The fourth door is editorial habit. When a site has no dedicated editor for each vertical, the person posting tags by feel. The feel says game means esports. The tag is applied, and the error enters the data.

Those four doors open at once, and the result is a photo set sitting next to a suspension.

What interests me is not a single error. A single error can be fixed in three minutes. What interests me is that this error replicates itself. Every time a piece like this gets tagged esports, the system relearns that content looking like this belongs in esports. Next time it tags itself with no human decision. After a year, a site's esports section can hold thirty percent content where nobody competes.

When an esports section holds thirty percent content where nobody competes, it is no longer a section — it is a keyword container.

And the price of a keyword container is not paid in ad money. It is paid in trust. A reader opens an esports section to find out who won. If that reader meets a photo set, they learn the section is not worth opening. Next time they open another source. And that other source may be abroad.

To understand why the photo set exists, you have to look at its actual value chain. The chain starts at the publisher. Manjuu and Yongshi design a character with features chosen for spread: contrasting colors, a silhouette recognizable at small size, a memorable detail in the rabbit ears. The publisher drops that character into the game through a banner. Players spend money to get the character. Some of those players make content: drawings, stories, embodiment photos. Among them, a few are professional cosplayers with skills in set building, lighting, costume, and post-production.

That cosplayer produces a photo set. The set needs somewhere to publish. The site needs high-click content. The two meet. Sometimes it is a paid placement relationship, sometimes a barter one — the cosplayer gets exposure, the site gets traffic. Nobody has to violate ethics. Both sides simply benefit.

In that chain there is no player, no coach, no competition contract, no standings table. There is a publisher, a character, a content creator, and a site. That is a marketing chain, not a competition chain.

If I place the two chains side by side, the difference lies in who carries risk. In a competition chain, risk sits with the player. A botched play can cost someone a starting slot, a contract, a year of a career. That risk creates drama, and that drama is what audiences pay to watch. In a marketing chain, risk is close to zero. A beautiful set gets praised, a less beautiful set gets feedback, and nobody loses a starting slot. A section that mixes two different risk chains teaches its audience that no risk is worth caring about.

I tested this hypothesis by looking at how large outlets split their sections. ESPN has a dedicated esports page and a dedicated gaming-culture page. Those two pages never merge their feeds. The reason is not academic snobbery. The reason is that readers of one page do not want notifications from the other. That separation does not reduce traffic. It increases it, because each page retains a loyal readership.

A small Vietnamese site does not have the budget to build two newsrooms. That is true. But a tag filter does not need a budget. It needs one person to sit down, read the definition of each section, and write one sentence per section. The esports section is where people compete and results exist. The gaming section is where characters and updates live. One sentence like that, taped to a wall, fixes most of the problem.

But wait. I have to talk about what I consider the most interesting part of this story, and it is not the error.

Inside the recommendation block of the photo-set piece, there is a real story. A Vietnamese PUBG player named Himass faces a possible suspension, tied to a community dispute and an apology from the organizing side. That is genuine competition content. There is risk in it, contracts, rules, professional consequences. If I want to know whether a site's esports section is real, I look at how it handles a story like that. Does it name the deciding body correctly? Does it distinguish allegation from ruling? Does it give a specific suspension length, cite which clause was violated, source the organizer?

A site that only runs photo sets and cannot run a story like that has a competence problem, not a tagging problem. A site that runs both has a classification problem, and that problem is far lighter.

I say this because I do not want my piece read as an attack on the cosplayer. The person behind that set did their job. They picked a highly recognizable character, built a set, shot it, published it. In the content economy, that is skilled, valuable labor. The article's caption describes the set as a rather impressive embodiment, with a mischievous spirit and a look close to the in-game version. I have no data to dispute or confirm that, because the article publishes no engagement metrics at all — no views, no likes, no shares.

And that is the second point I want to make clear. An article that says a product is impressive while offering no number to compare against is an advertisement, not an assessment. I do not blame the writer. I blame the template. In seven years of reading pieces like this, I have never once seen one that called a photo set mediocre. If every set is impressive, the word impressive means nothing.

Praise without a measuring stick is only a gift, and a gift cannot be used as data.

Now I want to return to the question I raised at the start. If I call this an error, whom am I protecting?

I am protecting the reader. More specifically, I am protecting a very vulnerable group of readers in the Vietnamese market: people newly entering esports and trying to learn how to read it. When a newcomer opens an esports section, they lack the background to filter on their own. They trust the label. If the label says esports, they learn that esports includes character photo sets. Six months later, they hold a distorted definition of the industry they are trying to understand.

That is real damage, and it does not show up on a traffic dashboard.

I also have to raise another possibility, one that makes me less certain. Maybe this site is doing it right, and the one who is wrong is me.

The argument against me is clear, and I want to present it seriously rather than build a straw man to knock down.

First, the line between game players and esports viewers barely exists in practice. A person can top up for an Azur Lane banner and follow PUBG Asia Stars. If that is the same person, putting both content types in one section serves exactly that person, and my separation is just an unnecessary academic habit.

Second, keeping a pure competition section can be a form of arrogance. It assumes only competition deserves reading. But the bulk of the gaming industry's revenue comes from players who never compete, and most of their time goes to collecting, looking at characters, and discussing design. If I call that part noise, I am dismissing the majority of the audience.

Third, and this is the strongest argument, character content may be funding competition content. If a photo set pays the bills, then the suspension story is written with the photo set's money. In a market where competition news struggles to earn directly, using easily viral content to subsidize hard-to-viral content may be a reasonable strategy, even the only viable one.

I weighed those three arguments. I weighed them again after finishing the original article. And I hold my position, for one specific reason.

The third argument describes a subsidy strategy. Subsidy is a valid strategy. But a subsidy needs a wall. If the photo set's money flows to the suspension story, the suspension story still has to sit in its own section, and the photo set still has to sit in its own. When the wall disappears, the subsidy becomes a substitution. Readers stop reading the suspension story, because the photo set has taken the screen space. That is the point where the third argument defeats itself.

The first and second arguments are correct, but they are correct about audience behavior, not about a newsroom's obligation. A site has the right to serve an audience with many interests. A site has the obligation to tell that audience what it is reading. Those two things do not exclude each other. They only exclude each other when the people making content are too lazy to label.

And here I have to admit something about myself. I have made this mistake. At nineteen, I wrote a piece about a transfer and tagged it tactical analysis, while half of it was just a story about a player I liked. An editor called me and asked one question: which paragraph holds the tactics. It took me twenty minutes to answer, and in those twenty minutes I knew I had no answer. I fixed the piece. I still remember the feeling, and I wrote this piece partly because of it.

There is a test I still use, and I want to offer it here as a tool rather than advice. Take any piece in an esports section. Ask: if this had been published ten years ago, would it still be true? A story about a sanction stops being true when the sanction ends. A story about a patch stops being true when the next patch lands. A cosplay photo set stays true, because it does not depend on timing. That durability sounds like a strength, but in a news section it is the signature of something that does not belong there. News has a shelf life. Content with no shelf life belongs in another section, and another section has another way of making money.

Now, the predictions. I offer three verifiable ones, and I will grade myself in six months.

One, Vietnamese esports sites will not solve the classification problem by building more newsrooms, but by moving part of their character content to social channels and keeping the main section for competition news. Signal to watch: the appearance of a separate section label for gaming and gaming culture on the homepage, split from the esports label.

Two, the Himass story and similar disputes will become the competence test for Vietnamese sites this year. Signal to watch: whether coverage names the deciding body, gives a specific duration, and cites the violated clause, or simply retells social media posts.

Three, the frequency of character pieces inside esports sections will not fall, because the traffic incentive will not fall. Signal to watch: the share of content with nobody competing among the latest forty headlines. My count this morning was twenty-one out of forty. I will count again at year's end.

When football stopped in March 2026, I was seventeen, and I made a podcast series about lower-division players nobody covered. I learned one thing that season and still carry it: people do not abandon a section because it is dry. They abandon it because it stopped telling the truth.

The Shimakaze photo set will still be there next week. It will still be beautiful. It will still have nobody competing. And the question I leave behind is not whether the set deserves to exist. The question is this: when a reader opens your esports section for the first time, do they learn that esports is a sport.