The KRAFTON Ledger: 9% Versus 100%, a 2.5 Million Won Fine, and Two Permanent Bans

In 2026, I opened a second-hand laptop in Guwahati and let 312 shots become a language. The reason was simple: the scoreline told me what happened, shot quality told me how. That habit taught me something I still use: to read an institution, you do not read its press release, you read its timestamps. Last year a different kind of scoreline landed in front of me. A screenshot sent by a kid in Sylhet. A fifth Premium Bundle, opened. The package text said that after four consecutive failures, the fifth attempt guaranteed the item. The grid came back empty. He opened a sixth, then a seventh. I saved that screenshot, because the empty grid is the most data-dense point of that week. A promised 100% against a real 9%: that is the calculation below. And the calculation is not only about a loot box. The sample: four incidents, one ledger From the source material I separated four incidents in chronological order. In 2026, ZDNet Korea reported that KRAFTON, for the first time in its IPO registration, disclosed that it provides technical services to Tencent-operated Peacekeeper Elite and receives service fees for them. In 2026, two separate matters inside the NewJeans collaboration: inaccurate odds information on paid bundles, and misuse of cosmetic skins. In June 2026, Yonhap reported that South Korea's Fair Trade Commission fined KRAFTON 2.5 million won for misleading loot box odds information. And on September 23, 2026, after a stream-sniping investigation at PUBG Asia Stars, two Vietnamese professionals received permanent bans. I do not want to read these as scandals. Scandal is an emotional category. I read them as a liability ledger, with four columns. Latency: how many days from incident to sanction. Duration: how long the sanction lasts. Cost: what the sanction is worth in money. Transparency: whether the definition behind the sanction was published in advance. The method is not new. PPDA was never a prophecy; it was a pressure map of Russia. In 2026 I logged PPDA for 64 matches, watched Germany's pressing collapse, and called nothing. It was a map showing where pressure builds and where the gaps open. Same here. I am not forecasting KRAFTON's future. I am showing where pressure accumulates inside its sanction architecture and where liability quietly disappears. The core: 91 percentage points between 100% and 9% The first layer is plain. During the NewJeans collaboration, two paid products entered the store: the PUBG X NewJeans Loot Pack and the PUBG X NewJeans Premium Bundle. Players opened them for collectibles and character customisation items. KRAFTON announced that if a player did not receive the set blueprint after four consecutive opens, the fifth attempt would deliver it at 100%. What actually happened is clear in the commission's file. Per the KFTC's finding, the real rate was only 9%. The gap between the announcement and reality: 91 percentage points. I want to keep two things separate here, because public discussion tends to merge them. One is a technical error: the Premium Bundle was in fact not covered by the bad-luck-prevention mechanism, and some in-game displays were inaccurate, which the PUBG team itself eventually confirmed. The other is the distance between a published rate and a real one, which the regulator examined separately. Put the numbers side by side and the picture changes. KRAFTON refunded roughly 1.1 billion won to about 380,000 buyers, and added in-game compensation worth around 9.8 billion won. The fine: 2.5 million won. I am doing the arithmetic myself here, because published reports rarely print the ratio. The fine against the in-game compensation package is 0.0255%. Divided across 380,000 affected buyers, the fine is about 6.6 won per person. The cash refund is about 2,895 won each. The in-game compensation is about 25,789 won each. Those three figures say one sentence. The ceiling of the regulator's punishment and the value returned to the consumer are both small, but they are not on the same scale. One is a few won per person. The other is a few thousand won of credit per person. And because the company corrected the information and paid compensation on its own initiative, the KFTC's action stayed limited to a financial penalty. That is the ledger's most important line: voluntary correction converts a structural sanction into a line-item cost. I am keeping latency approximate here, because the exact launch date of the bundles is not in my file; I only have 2026. Assuming the start of 2026, the run to June 2026 is close to a year. That is an estimate, not a measurement, and I will not seat an estimate where a measurement belongs. The core: harm nobody purchased The second matter is another face of the same collaboration. According to Korean media, some players took the NewJeans members' face skins and combined them with revealing outfits to create and circulate sexually harassing images and videos. It drew more attention because Haerin and Hyein were not yet adults at the time. KRAFTON and ADOR, NewJeans' management agency, said they would act against uses of the items outside the collaboration's original purpose. KRAFTON then restricted combining the NewJeans face skins with certain outfits. Debate followed, and this part interests me more. One camp asked a simple question: is the problem user behaviour, or is it PUBG's own customisation architecture? If the game lets you join a large number of slots at once, and face skins are exact human likenesses, then the route to misuse was built into the design. I admit I cannot count this section. How many combinable slots exist, how many outfit tags, what the defaults are: I do not have those numbers. Without numbers I will not write a measurement. What I can say is that of the four incidents, this is the only one where harm fell on people who had bought nothing. In the other three, players were consumers. And it was answered with a restriction, not with a corrected number. The core: how information arrived on a regulatory document The third incident is first in time, 2026. ZDNet Korea reported that in its IPO filing, KRAFTON disclosed for the first time that it provides technical services to Tencent-run Peacekeeper Elite and collects fees. The context already existed: in May 2026, PUBG Mobile shut down in China over licensing, after which Tencent moved players to Peacekeeper Elite, altered in gameplay, visuals and content. KRAFTON had claimed the two were separate products. I do not have the fee figures, so I will not rank this incident in the ledger. What I can count is the timing. The most financially sensitive detail did not arrive through a press release. It arrived through a securities filing. That is a pattern, not an accident. Information that investigators could reach anyway tends to reach announcements late; information that creates legal exposure if withheld goes into the document. For players the meaning is plain: outsiders never see the whole map of a deal. A press conference holds the part of the map that prints well. The core: the day the sanction took a few days The fourth incident is the most recent, and here the ledger bends the other way. At PUBG Asia Stars, Korean streamer Soopi of Gen.G accused two Vietnamese professionals, Himass of Anyone's Legend and TanVuu of The Expendables, of watching opponents' live streams. Organisers immediately removed both from the rest of the tournament, adjusted points, increased stream delay, cancelled the third match day and split the prize money. On September 23, KRAFTON published its investigation. Its conclusion: both players used information from outside, including other players' live streams, to make reads and build tactics. It called this stream sniping, a violation of the PUBG: Battlegrounds operating policy and the professional conduct rules. The outcome: permanent account bans, plus exclusion from all official PUBG Esports events organised or approved by KRAFTON, including PGC, PGS and PNC. The investigation found no further violations. Three ledger columns fill here. Latency: days, not even weeks. Duration: infinite, with no expiry date announced. Cost: a career. Set against the 2.5 million won fine of June 2026, a time asymmetry becomes visible. Sanctions against players arrive in days and last indefinitely. Sanctions against the regulator's subject arrive in years and last one line of a calculation. The reaction is also countable. An appeal carrying more than 4.1 million signatures. Several well-known Vietnamese community figures — Độ Mixi, PewPew, Rambo, DjChip and Ngân Sát Thủ — protested or removed PUBG. GAM x The Expendables demanded a clear explanation, and Anyone's Legend stood up for its player's right to be heard. The contrarian angle: speed and transparency are not the same thing The easy reading is that the company is soft on itself and harsh on players. I tested that explanation first, because it is the comfortable one. There is an alternative: the two cases may not be comparable at all. One is a consumer-protection matter with a legal ceiling on the penalty; the other is a competitive-integrity matter, where the standard esports remedy is a permanent ban. I am not dismissing the alternative, because it is partly true. But something survives even after granting it, and that is the real blind spot. The problem is not severity. The problem is latency and the transparency of the definition. The wrong number that took money from players was written by the publisher itself, and correcting it required community pressure plus an investigation that ran into 2026. On the other side, permanent bans arrived within days — but under a rule whose boundary no player can verify from inside a match. How many seconds of stream delay equal information from outside is not a calculation anyone in the lobby can perform. That is my objection. Where a rule is measurable, the system is slow; where a rule is unmeasurable, the system is fast. Precision is cheap to enforce: you match a number and you are done. Ambiguity is expensive to define, because every definition risks losing a future case. Institutions choose conveniently which one they prefer to do. And 4.1 million signatures are not proof of correctness to me. I never treat a vote as a substitute for evidence. But 4.1 million signatures do prove something else: how many people believe the definition was never shown to them. When a community refuses to accept a sanction, that is not emotion. It is the output of a process. Where the definition is not printed in advance, every decision reads as arbitrary, even when the decision is fair. Takeaway: what to count next round Two signals are worth watching against this ledger. First: whether KRAFTON publishes a minimum tolerance threshold for stream delay or outside information before the next PGC. If it does, the definition becomes countable for the first time, and players at least know which boundary they are playing inside. Second: whether the principle that voluntary correction reduces the penalty becomes a recurring audit item for the KFTC in consumer-rate cases. If it does not, the ledger line I found most useful — the one saying the price of liability got cheaper for the regulator — will turn out to be wrong. The transfer window is a ledger, not a rumour mill. So are these four incidents. And reading the ledger turns up nothing but one pattern: corporate accountability is measured in years, player accountability in days, and the two are never placed on the same scale. A player knows what the end of a career looks like. Nobody has written down yet what community trust costs per won.