2.5 Million Won vs a Career: KRAFTON's Punishment Pyramid Is Standing Upside Down

2.5 Million Won vs a Career: KRAFTON's Punishment Pyramid Is Standing Upside Down

Hook

September 23, 2026. Within a day of KRAFTON publishing its investigation report, a number started circulating through Vietnam's PUBG community: 4.1 million signatures. One demand — justice for Himass of Anyone's Legend and TanVuu of The Expendables. KRAFTON had said the two players used outside information, specifically opponents' livestreams, to make calls. Everyone recognised the term: stream sniping. The result was familiar too: permanent account bans and exclusion from every official PUBG Esports event run or approved by KRAFTON, including PGC, PGS and PNC.

A few months earlier, in June 2026, a Yonhap report carried another number that many scrolled past. South Korea's Fair Trade Commission fined KRAFTON 2.5 million won for misleading odds information in PUBG: Battlegrounds. How misleading? The announced guarantee said four failed attempts would make the fifth a certain 100% for a set blueprint. The actual rate was 9%. KRAFTON later refunded roughly 1.1 billion won to about 380,000 buyers and issued in-game compensation worth around 9.8 billion won.

On one side, 2.5 million won. On the other, two careers. In KRAFTON's punishment pyramid, a player's integrity costs a career and a publisher's information costs a rounding error. This piece tries to measure that pyramid.

Context: a trust ledger for an industry

In 2026, sitting in a PUBG Mobile casting booth in Bangladesh, I learned that South Asian communities judge a publisher by the legitimacy of decisions, not by the code base. Chat never asked about patch notes on the website. It asked why the rule that made a buff possible changed today and would not change tomorrow. Back then it felt like small-community restlessness. Today it reads like an unwritten contract with the publisher whose currency is not money but trust. KRAFTON has broken several articles of that contract.

Article one, the 2026 NewJeans collaboration. PUBG X NewJeans Loot Pack and Premium Bundle — two paid products carrying collectible skins inspired by the members. The problem began with information: players said they bought more than five Premium Bundles without receiving the set blueprint. The PUBG team later conceded that the Premium Bundle was outside the bad-luck prevention mechanism and that some in-game displays were inaccurate.

Article two, the ethical spillover. Some players combined NewJeans face skins with revealing outfits to produce and circulate sexualised images and videos. The matter grew heavier because Haerin and Hyein were minors at the time. KRAFTON and ADOR said action would follow, and certain face-skin and outfit combinations were restricted.

Article three, the 2026 China file. ZDNet Korea reported that in its IPO securities filing, KRAFTON publicly acknowledged for the first time that it provides technical services to Tencent-operated Peacekeeper Elite in China and receives service fees. PUBG Mobile shut down in China in May 2026 over licensing issues, and Tencent moved players to Peacekeeper Elite, a title with altered gameplay, visuals and content.

Article four, September's Asia Stars. After Gen.G streamer Soopi's accusation, organisers removed Himass and TanVuu from the rest of the event, adjusted scores, increased livestream delay, cancelled the third day and split prize money equally.

Core: whose gap, whose bill

The loot box case is mathematical, which is why it is the clearest. When randomness is the entire value of a product, the odds disclosure is not marketing material — it is the product specification. A wrong specification means a wrong product. For a player, 100% means 100%, not 9%. This could have been a typo if the fact that Premium Bundle sat outside the mechanism had been written anywhere. Had it been, nobody would have bought more than five. Publishers prefer opacity because loot box economics stand on it. But when the published number is wrong, opacity becomes strategy, and strategy has a legal name.

Here is my second calculation. If compensation is roughly 3,900 times the fine, the fine is not a punishment — it is a receipt for a discount. The FTC itself said KRAFTON corrected the information and compensated buyers, which is why the matter was limited to a financial penalty. The price of the offence was set by how fast you can apologise.

My third calculation is where my suspicion lives. In the NewJeans face-skin case, KRAFTON's first fix was banning combinations. But the problem does not end in user behaviour; it starts in system design. If character customisation never respects a boundary, and a performer's face becomes a modding surface, blocking specific outfits is putting a chair in front of an open door. Something bigger shows up here: neither brand in a brand deal actually controls its risk. Licence an idol's image and you also licence the user's imagination, and nobody prices that discount upfront.

The China account is simpler. Without the IPO filing we would not have known. When a publisher conceals the technical relationship between its flagship China product and another title until securities law forces disclosure, the community's reading is not unfair: here transparency is not policy, it is obligation.

Now the September decision. KRAFTON said its investigation found no further violations among others. Many read that as a clean certificate. It is a technical statement: the investigation did not find what the investigation set out to find. Not an acquittal, but an admission of scope.

And the clock? Stoppage Time Rajshahi started as noise; after 4.1 million signatures in Vietnam, it became the only clock I trusted. Those signatures are not a legal instrument. They are a pressure clock. That is KRAFTON's real problem: however reasonable the reasoning, if the scale of punishment is not written down, the 4.1 million people outside the announcement will read every decision as political.

One thread ties all four episodes: in each, KRAFTON's first reflex was a communication fix, not a code fix. The tooltip was corrected, the system was not.

Contrarian: how I could be wrong

I am not assuming everything cuts one way. Take the other side.

First objection: in esports, stream sniping is the one offence that erases the game itself. If opponents know their screen is being watched, positioning training means nothing and the whole competition becomes theatre. KRAFTON's operating policy and professional conduct rules say exactly that, and a lifetime ban is proportionate, not cruel.

Second objection: if I use those 2.5 million won as the shame benchmark, my own arithmetic collapses. Korean law caps FTC civil penalties in a narrow window, and voluntary compensation constrains how the penalty is valued. Read that way, the real cost is the 9.8 billion won compensation. In that mirror, the system worked.

Third objection, the strongest against my own take: 4.1 million signatures are not a court. This is a movement powered by national identity and anger. It asked the right question in the wrong format. The demand should have been: publish an appeals protocol, publish a written disciplinary scale, publish a capability statement for investigations. Argument belongs to the regulator; slogans belong to the stage.

Fourth, against myself again: on the NewJeans face-skin story, many in the community argued the fault lay in the game's customisation system rather than individual conduct. I first read that as evasion. Now it looks like a claim worth standing on.

The rooftop gave me the take, the fall gave me the context. After Germany's loss to Mexico in 2026 I called the back three a museum. The take was right, the reason was wrong — a back three is not bad, it simply was not working then. Here too the first read holds: KRAFTON has spent the community's trust. The reason is different — this is an operations company, and an operations company's reflex is public relations, not product repair.

Takeaway: a falsifiable bet

Here is my bet, and I will check it against the clock: within twelve months, PUBG Esports will publish a written, tiered disciplinary scale and an independently audited odds-disclosure statement. The reputational cost of ad-hoc punishment has now exceeded the cost of punishment itself. If it does not, expect at least one more boycott wave across South and Southeast Asia.

And the question on our side: PUBG Mobile is the largest existing screen sport in Bangladesh. When trust is genuinely a cost line, how much of our name is written in KRAFTON's trust ledger, and how often do we sit at the table where the rules are drafted? If we do not keep the accounts, the bill lands on our card.