PUBG's Nine Percent: KRAFTON's Four Scandals and a Single Structural Fault

It was nearly three in the morning at a gaming cafe in Colombo. Plastic chairs on the floor, PUBG Asia Stars 2026 on the projector, sixteen phone screens glowing on the table. When the news of the cancelled third match day hit the screen, I was not really watching gameplay. I was watching the chat. And what was happening in that chat was not an argument about map rotation. It was a live accounting of trust being withdrawn from an institution. Since that night, one suspicion has hardened into the central claim of this piece: PUBG's biggest problem is not the lid of the loot box, it is the monopoly on decision-making -- where the publisher, the organiser and the judge are the same company. On September 23, KRAFTON announced that Himass of Anyone's Legend and TanVuu of The Expendables had used external information, specifically opponents' livestreams, to shape in-game decisions. In the company's language, stream sniping. The penalty: permanent account closure and a ban from every official PUBG Esports event run or approved by KRAFTON, including PGC, PGS and PNC. The accusation came from an opposing streamer. The verdict came in a single step, at maximum severity.

On the night Boeser made my bad take age in real time, I was sitting at a bar table in Vancouver, watching a claim of mine dissolve live. That night taught me that the most dangerous thing in a live moment is a sentence spoken without a timestamp. So let me set this down plainly: this article is written nine days after the ruling, and the meta of that ruling has not settled. The empty NBA bubble taught me that silence can still boo a bad take. Here there is no silence -- there are 4.1 million signatures. The question is therefore not whether Himass is guilty. The question is what happens when the same company admits fault on four different issues in five years: its judicial architecture becomes the thing under review.

Context: one company, three roles

Inside the PUBG: Battlegrounds ecosystem, KRAFTON's role differs from any major sports league. In football, the referee, the league authority and the club owners are three separate entities. In PUBG, the owner of the game, the organiser of the tournament and the disciplinary committee sit under one corporate roof. That integration buys efficiency, but it creates one unavoidable question in every dispute: who hears the appeal?

Across 2026 to 2026, at least four episodes have put KRAFTON in front of that same question. In 2026, the admission in its IPO filing that it provides technical services for Tencent-operated Peacekeeper Elite in China. In 2026, incorrect probability information on items from the NewJeans collaboration. In the same year, the harassment controversy around the face skins from that collaboration. And in 2026, the stream sniping ruling at PUBG Asia Stars. Four incidents, four different categories, yet the closing answer was identical each time: we are fixing it, we apologise, and we are refunding what happened.

This is where the resemblance to football's VAR argument shows up. The controversy migrates off the pitch and settles in the review room and the grey areas of the rulebook. The distance between the spectator and the decision does not shrink, it grows. The same thing is happening in PUBG. Before buying an item, the question was what the item looks like. Now the question is whether the fifth attempt was genuinely guaranteed. During a match, the question was who played better. Now the question is whose stream was open on whose screen.

The market does not allow these questions to be treated as small. Vietnam, Bangladesh, India, Sri Lanka -- the South and Southeast Asian PUBG community is simultaneously the publisher's largest market and its loudest mirror. In 2026, casting on the Dhaka PUBG Mobile circuit under the handle TimeBurner, nearly every tournament day I covered produced a rule dispute, and every one of them was settled by a decision spoken aloud, with no published penalty table. I did not understand then that this was the miniature version of the cases to come.

Core analysis: what the nine percent actually says

In the 2026 NewJeans collaboration, KRAFTON placed two paid products in the in-game store: the PUBG X NewJeans Loot Pack and the PUBG X NewJeans Premium Bundle. The marketing stated that if a player failed four consecutive times to obtain the set blueprint, the fifth attempt would deliver it with 100 percent certainty. Then came the complaints: many players bought more than five Premium Bundles and never received the item. The company later confirmed that the Premium Bundle did not in fact fall under the pity-prevention mechanism, and that some in-game information displays were incorrect. South Korea's Fair Trade Commission later published the number: the real rate was 9 percent.

That 9 percent figure is the single largest piece of information gain in the whole case. The loot box economy does not sell items. It sells a narrative of near-certainty. The so-called pity system makes a random purchase look as plannable as an instalment plan; which means a broken promise is not a bug, it is a broken contract. The gap between 100 and 9 is not a gap in probability maths. It is a gap in the basis of pricing. If the chance on the second attempt is 9 percent while the buyer believes the fifth attempt is 100 percent, is that price being paid voluntarily?

Put the money side by side and the picture sharpens. According to Yonhap, in June 2026 the FTC fined KRAFTON 2.5 million won. The company refunded roughly 1.1 billion won to about 380,000 buyers, and issued in-game compensation worth roughly 9.8 billion won. Divided out, that is an average refund of about 2,900 won per buyer and around 25,800 won in compensation value. The fine works out at roughly one four-thousand-three-hundredth of the total compensation. The number says the penalty was never the price of the deception. The refund and the compensation were the cost of rebuilding goodwill. Because the company voluntarily corrected the information and paid compensation, the sanction was limited to a fine.

The flaw in that model is that it looks like cooperation to the regulator while teaching users that getting caught means getting refunded. The cost that lands in the company's books as goodwill recovery becomes, for the user, an insurance premium against future risk. And then comes the second question: who verifies the probability numbers? If the figure shown inside the game comes from the publisher's own system, the only route to proving a gap between that number and lived player experience is community sampling plus an authority's investigation. Against that backdrop, the FTC's discovery of the 9 percent is a rare public measurement.

The second blow from the same collaboration arrived from an entirely different direction. Korean media reported that some users combined NewJeans members' face skins with revealing outfits to produce and circulate sexualised images and videos. The sensitivity deepened because Haerin and Hyein were minors at the time. KRAFTON and the management company ADOR said measures would be taken against uses outside the collaboration's original purpose; KRAFTON then restricted combining the NewJeans face skins with certain in-game outfits.

The response framed the problem, almost perfectly, as user misconduct rather than product design. A large part of the community asked the reverse question: if the boundary of customisation can be broken that easily from outside, why does the responsibility sit with the user alone? The economic decision to turn teenage idols' faces into usable skins was the company's. Here too the same convenient solution appears -- impose a limit, do not change the sales model. Refunds and bans are cheap; rebuilding the store is expensive.

The third episode is the largest geographically and the least discussed. In 2026, ZDNet Korea reported that KRAFTON had disclosed for the first time, in its pre-IPO registration filing, that it provides technical services to Tencent-operated Peacekeeper Elite in China and receives service fees for them. The timing matters: in May 2026, PUBG Mobile shut down in China after licensing problems, and Tencent moved players to Peacekeeper Elite, a game with significant similarities but altered gameplay, visuals and content. KRAFTON had previously asserted the two were separate products.

In this case no one banned an account and no one apologised, yet it is the most enduring of the four. Because it is an admission of institutional interest: when one brand runs under two names in two countries, the weight of the separate products claim rests on legal definition rather than on the game itself. The 2026 disclosure did not change the game; it changed who owns that game's life in China, as a matter of definition.

Now to the episode that explains the night this article opened with. The stream sniping accusation at PUBG Asia Stars 2026 was first raised by Soopi, a Korean streamer for Gen.G. The organiser's steps were fast and contested: Himass and TanVuu removed from the rest of the tournament, points adjusted, livestream delay increased, the third match day cancelled, and prize money split evenly. The September 23 verdict found that the two players had used external information, especially opponents' livestreams, violating both the PUBG: Battlegrounds operating policy and the conduct rules for professional players. Permanent account ban, and exclusion from PGC, PGS and PNC. KRAFTON said no further violations were found in other cases.

I have no objection to the finding of guilt. Using spectator information to shape in-match decisions breaks the core contract of the sport. My objection is the absence of a penalty ladder. In sporting governance, a first offence, deliberate cheating and carelessness normally get separate rungs: a hearing, a statement of reasons, an appeal, then the sanction. Here the sequence was accusation, mid-tournament removal, a later investigation, then the maximum penalty. When a first-time, information-based offence draws an immediate lifetime ban, the distinction between deliberate cheating and negligence is erased -- and the 4.1 million signature petition is in fact a protest against that erasure.

To understand why the wound runs deep, the number is enough. 4.1 million signatures equals roughly four percent of Vietnam's population. And that voice did not stay in chat. Parts of the player base, among them Độ Mixi, PewPew, Rambo, DjChip and Ngân Sát Thủ, either objected publicly or removed PUBG. GAM x The Expendables demanded a clear explanation, and Anyone's Legend spoke up for its player's right to be heard. This is not merely fan fury. It is a market signal: in the region that supplies KRAFTON with its highest active user base, a spread of doubt about the fairness of the judicial process costs more than a tournament prize pool.

This is where the real information gain sits, once the four episodes are placed side by side. The shared pattern is not loot boxes, skins, licensing or stream sniping. The shared pattern is that the remedy always arrived after the incident, always took the form of money or a ban, and always lacked a rule published in advance. KRAFTON's problem is not corruption; the problem is that the rule is written after the incident -- and a rule written afterwards never commands as much trust as the verdict commands speed. That is why the clearer the refund arithmetic becomes, the louder the question returns: who writes the rule next time?

How I could be wrong

I know my own trap, so let me concede it up front. First, I have welded four separate incidents into one structural thesis, and my familiar weakness is turning every small event into evidence for a big narrative. The four differ in timing, category and chain of decision: one licensing dispute, one consumer rights case, one content moderation case, one competitive discipline case. Bringing them under one umbrella means taking the benefit of pattern-hunting, and that benefit carries a price.

Second, my objection to the stream sniping penalty may be nothing more than comfortable sympathy. In international sport, doping operates on strict liability -- intent does not need to be proven at all. If competitive integrity is a one-strike regime, then the maximum penalty on the first offence is not an injustice, it is the rule. There is specific evidence against Himass and TanVuu, and I have no basis in hand to presume them innocent.

Third, I have measured the mood of Korean and South Asian forums through the volume of Vietnamese chat. Part of the Korean community reads this ruling as ordinary competitive order, and that side got too little room in my story. This piece is therefore a provisional claim written after a ten-minute cooldown. If KRAFTON stands up a published penalty table and an independent appeals panel within the next two quarters, my central thesis dies -- and that would be good news for the game. A transfer rumour is a campfire story until someone packs a suitcase. Here too, watch the institution, not the announcement.

Not a conclusion, but the next three dates

Back in Vancouver I used to chase the roar, until I learned to listen for the click, because the click does not happen live -- it happens on the replay. So this piece ends with three verifiable dates rather than a summary. One: does KRAFTON announce a published penalty table or an appeals panel before PGC? Two: do the store's probability figures come under third-party verification? Three: within the next eighteen months, does a similar probability dispute end in a second round of refunds?

In a game where one company owns the pitch, the whistle and the replay, who referees the referee?