Nine Per Cent Against One Hundred: Four Entries in KRAFTON's Trust Ledger

There is a number in the Korea Fair Trade Commission's ruling that does not catch the eye on a first read: 9 per cent. In June 2026, a Yonhap report confirmed that the commission had fined KRAFTON 2.5 million won for supplying misleading item drop-rate information inside PUBG: Battlegrounds. The announcement inside the game had said something else entirely — fail four times in succession and the fifth attempt guarantees the blueprint set, a rate of 100 per cent. The gap between the announcement and the record is 91 points. I mark that kind of gap as second-hand evidence, because it is not a rounding error; it is the arithmetic collapse of a promise. Before I put the phone down, I remembered 2026 — I opened a second-hand laptop and let 312 shots become a language. That language taught me my first rule: when the announcement and the record disagree, believe the record. Here, again, I had to. KRAFTON is a South Korean publisher whose main asset is PUBG: Battlegrounds, a battle royale that arrived in early access in 2026 and puts roughly a hundred players into every match. Its esports ecosystem runs championships such as PGC, PGS and PNC, and around them an enormous player and viewer base across the Asia-Pacific. In 2026, South Korea amended its games industry rules to make probability disclosure mandatory for paid items. The logic behind the amendment was simple: buying a loot box is an economic decision, and an economic decision requires the number. Against that backdrop, several recent KRAFTON episodes need to be read together. One is the 2026 NewJeans collaboration and its drop-rate misinformation. The second is the harassment that grew out of that same collaboration's face skins. The third is the disclosure that the company provided technical services to Peacekeeper Elite in China. The fourth is the 2026 PUBG Asia Stars, where two Vietnamese professionals were accused of stream sniping and sanctioned. The four episodes speak different languages, but one question runs through all of them: who keeps the accounts when published information and operative information diverge? Take the 2026 collaboration first. KRAFTON placed two paid products in the in-game store — the PUBG X NewJeans Loot Pack and the PUBG X NewJeans Premium Bundle. Players opened them for collectible items and character customisation. The trouble began in the drop-rate disclosure. The announcement stated that after four consecutive failures the fifth attempt guaranteed a blueprint set, at 100 per cent. Numerous players reported buying more than five Premium Bundles and still not receiving the item. After community feedback, the PUBG team acknowledged that the Premium Bundle did not actually sit inside the pity-mechanic structure, and that some in-game displays were inaccurate. KRAFTON corrected the text, apologised and announced compensation. In June 2026 the commission stated that the real rate was 9 per cent while the announcement promised certainty. The company refunded roughly 1.1 billion won to about 380,000 buyers and issued in-game compensation worth about 9.8 billion won. Because it corrected the information and compensated on its own initiative, the penalty stayed limited to a fine. This is where the accounting needs to get specific. I put the calculator down on the table. 1.1 billion won divided by 380,000 buyers is about 2,895 won per buyer in cash. 9.8 billion won divided by the same figure is about 25,789 won each — but in in-game currency, not in a bank. The two compensation ledgers are not the same ledger. One has a written value; the other has little beyond book value. The fine of 2.5 million won is roughly 0.23 per cent of the cash refunds and 0.026 per cent of the in-game compensation. Put those three numbers side by side and you see that the regulator's instrument and the company's apology economy operate on different scales. Discussion usually fixes on the size of the fine, but the buyer who purchased five bundles and gave up on the third does not appear fully in the cash-refund column. To me this is not a moral verdict; it is a gap in the accounts. I should be explicit: I have not independently verified the 9 per cent figure, and the reports do not detail how many trials the recorded ceiling was measured across. That uncertainty stays outside my arithmetic, not inside it. The second episode from the same collaboration sits on a different layer. South Korean media reported that some players used the face skins of NewJeans members to create and circulate sexually harassing images and videos, combining those faces with revealing outfits. The case drew further attention because Haerin and Hyein were not yet adults at the time. KRAFTON and ADOR, the group's management company, said they would act against uses of the items that fell outside the collaboration's original purpose, and KRAFTON then restricted combining NewJeans face skins with certain outfits. That decision also stirred debate. One section of players argued the problem lay in PUBG's own character design and customisation, not only in user behaviour. Before entering that argument I note an uncomfortable omission: no patch note separately states how many combinations the restriction severed and how many it left intact. Without a version note, the boundary of a rule cannot be measured — nor can who was protected where. The China question is old, but its documentary source is clear. In 2026, ZDNet Korea reported that, in the filing submitted ahead of its IPO, KRAFTON publicly acknowledged for the first time that it provides technical services to Tencent-operated Peacekeeper Elite and receives a service fee for it. In May 2026, PUBG Mobile shut down in China over licensing problems; Tencent then moved players to Peacekeeper Elite, a version altered in gameplay, visuals and content. KRAFTON's position has always been that the two are separate products. The question remains for players and observers: if the technical spine of one game is the same and one replaced the other, which metric stands as evidence for the claim of separateness? On such questions I avoid speculative comment, because the internal code and the contract's limits are not in my hands. I count what is on the record. The most recent episode came at PUBG Asia Stars 2026, and it involved two Vietnamese players — Himass of Anyone's Legend and TanVuu of The Expendables. On 23 September, KRAFTON published its investigation findings, stating that both had used outside information, including other players' livestreams, to reach judgements and build strategy during matches. KRAFTON classified this as stream sniping and said it violated both the PUBG: Battlegrounds operating policy and the code of conduct for professional players. The punishment was a permanent account lock and a ban from all official PUBG Esports events organised or approved by KRAFTON, including PGC, PGS and PNC. The affair began with an accusation from Gen.G's Korean streamer Soopi, who said the two Vietnamese players were watching opponents' streams. Organisers then removed Himass and TanVuu from the rest of the tournament, adjusted the points, increased livestream delay, cancelled the third match day and redistributed the prize money. After the investigation, KRAFTON said no further violations were found in other cases. The speed of the sanction produced a strong reaction in Vietnam's PUBG community. Some 4.1 million signatures were collected demanding justice for Himass and TanVuu. Độ Mixi, PewPew, Rambo, DjChip and Ngân Sát Thủ either protested or removed PUBG. GAM x The Expendables demanded a clear explanation from the organisers, while Anyone's Legend argued for their player's right to be heard. I reconcile the timestamp before I let the headline breathe, and here the crucial point is this: the announcement delivered a decision, but the chain of evidence leading to it — which timestamps, which stream, how many seconds of delay, how they were matched — was not published in verifiable form. The delay was increased, so the match timeline itself changed; somebody holds the count of that changed rule, and it is not me. In a quiet room, keeping count of sound matters more — Guwahati taught me that a quiet room can hold a whole league. Now to the counter-intuitive part, because my own trade has handed me one warning reference from football. At the 2026 World Cup in Russia I logged PPDA and flagged Germany's pressing collapse in advance. PPDA was not a prophecy; it was a pressure map. But where PPDA counts pressing events, esports has no direct equivalent; the comparable measure here is event density per round, information lag, and the rate of utility exchange. Without that translation table, dragging a football metric into esports makes the number decoration rather than evidence. So I will not weld the four episodes into one story. Structurally they differ: one is a measurable misstatement, one a design and moderation gap, one a corporate-structure ambiguity, one a crisis of adjudication timing. Fuse them and the demand turns foggy — it sounds like 'we want justice' but cannot be checked. What can be checked is a list: correction dates, rule version numbers, and a record of comparable past sanctions. Some want to merge the four into a single number called 'distrust in the publisher'. I cannot do that. Correlation is not causation. I will also say this: that Himass and TanVuu are innocent is a claim I lack data for; that they are guilty is equally beyond me. My rule is two-tournament confirmation, never a verdict from a single sample. The rule that kept Damsgaard's file open in 2026 keeps this file open too. What will I watch in the next round? First, whether KRAFTON publishes a public precedent ledger for esports sanctions — a checkable list of how many days, how much money and what scope of ban comparable offences drew before. Second, whether the compensation formula after the drop-rate correction becomes a template for other publishers, and where the ratio between cash refunds and in-game currency settles. Third, whether Korea's probability-disclosure model travels to other markets. The transfer window is a ledger, not a rumour mill — and the most important column of this ledger has still not been opened for inspection: how many buyers, having purchased five bundles, were lost while still trying to do the sum.