The Label Said 100%, the Code Said 9%: The Price of Trust Across Three KRAFTON Crises

In June 2026 I closed my laptop halfway through reading a Yonhap report. South Korea's Fair Trade Commission had ruled that KRAFTON must pay a fine of 2.5 million won for supplying misleading information about item drop rates in PUBG: Battlegrounds. The number looked so small that I first assumed a decimal had slipped. The next paragraph explained why the story sat there. In the same case, KRAFTON refunded roughly 1.1 billion won to about 380,000 buyers and issued in-game compensation worth roughly 9.8 billion won.

The two numbers at the center are these: the game stated that after four consecutive failed attempts, a player would be guaranteed the set blueprint on the fifth attempt, at a 100% rate. The commission calculated the real rate at 9%.

A company that prints 100% and delivers 9% is not selling a product. It is selling a promise. In 2026, standing on the terraces of Hongkou Stadium at the Shanghai derby, I learned that the scoreboard is the least reliable witness in the room. Shenhua lost 6-1 that day, the stands erupted, and I was watching a midfield pressing as if it were chasing a narrative rather than points. I later wrote a school column built on 18 shots and 62% possession. Everyone wrote the collapse; nobody wrote who kept running. Across these three KRAFTON crises I went back to that same place — not the outcome, the behavior.

Context: where an odds figure becomes part of the price

The logic of KRAFTON's item economy hides inside an old casino rule. When a player buys a paid loot box, they are not buying a specific item; they are buying a range of probabilities. A published drop rate is therefore not a marketing line, it is part of the price structure. What the industry calls a pity system reads as protection to the player and as a spending ceiling in the accountant's ledger. A regular player on a middle income draws their monthly budget precisely on that ceiling.

In 2026 KRAFTON launched a collaboration with the Korean girl group NewJeans. Two paid packs entered the store: the PUBG X NewJeans Loot Pack and the PUBG X NewJeans Premium Bundle. Opening them yielded collectible items and NewJeans-inspired customization. The controversy formed around the odds disclosure, which promised a guaranteed set blueprint on the fifth attempt after four failures.

Many players reported buying more than five Premium Bundles and still not receiving the item. After community pushback, the PUBG team acknowledged that the Premium Bundle was in fact not part of the bad-luck-prevention mechanism, and that some in-game display information had been incorrect. KRAFTON corrected the information, apologized and announced compensation. A large share of those who had spent heavily demanded refunds.

In Korea this falls under consumer protection law, where odds disclosure is regulated information. KRAFTON's error was not merely a wrong message; it was a wrong regulated figure. In June 2026 Yonhap reported that the company had refunded about 1.1 billion won to roughly 380,000 buyers and granted in-game currency compensation worth about 9.8 billion won. Because it voluntarily corrected the information and paid compensation, the sanction stayed limited to a fine.

After those numbers, PUBG's name is still attached to three more disputes — the NewJeans skin harassment case, the Peacekeeper Elite question in China, and the bans on two Vietnamese players at PUBG Asia Stars 2026. All three are faces of one question: how clearly does the platform write its rules down, and who carries the burden of interpretation.

Core analysis: the ceiling, the invoice and the silence

A pity system is not generosity; it is a pricing instrument. Rather than limiting the customer's risk, it fixes the publisher's revenue band. The arithmetic is simple. If an item's success rate is 9% and each draw is an independent event, a fifth attempt after four failures still carries the same 9%. But a player who sees 100% printed is entering a different contract, in which the fifth purchase means a certain outcome. When the disclosure is wrong, the contract's foundation breaks, and the cost of that break is not only money but time and trust.

After the story broke I ran a small test of my own, because I do not publish what I have not calculated on paper. Assume a box's success rate is 9% and a player has decided to buy no more than five. The probability of at least one success across five purchases lands near 38%, meaning roughly two out of every three players walk away empty-handed. The in-game message said 100%. The gap between those two numbers is what ends a regular player's monthly budget. Inside a loot box, the word 'guaranteed' is really a price ceiling; breaking the ceiling breaks the buyer's entire budget plan.

Run the compensation numbers once more. About 1.1 billion won went back to roughly 380,000 buyers, an average of about 2,895 won each. Beside it sits in-game compensation worth about 9.8 billion won. Combined, player losses absorbed by the company reach roughly 10.9 billion won, against a fine of 2.5 million won. The fine works out to roughly 0.02% of the total compensation — the regulatory signal is blunt: fast admission plus compensation is the cheapest defense available.

This is not a moral argument, it is an incentive argument. In a structure that rewards rapid correction, a publisher's smartest play becomes making the error first and owning it quickly afterward. From a market perspective that play is not cheap, it is deferred. Every such episode discounts the odds figure in players' minds. At the next event, the player does not believe, the player calculates. And in a community that calculates every purchase, the revenue curve for limited-time events slowly flattens out.

For comparison, China's cultural regulatory framework has required online games to publish random item drop rates since May 2026. That level of transparency is not technically difficult; it is a question of political will. A company that discloses to two markets under two different standards risks more than regulators — it risks the community's memory.

The second NewJeans dispute in 2026 was a different animal. Korean media reported that some players used face skins of NewJeans members to create and circulate sexualized and harassing images and videos, pairing the faces with revealing outfits. The case landed harder because Haerin and Hyein were not yet adults at the time. KRAFTON and ADOR, the group's management agency, said they would act against uses outside the collaboration's original purpose. KRAFTON then restricted combining NewJeans face skins with certain outfits.

The restriction points at the target. In a customization system where faces, outfits and poses lock together, the system itself owns the question of who may combine what. Some players immediately asked whether the problem sat in user behavior or in PUBG's design. When a platform locks outfit combinations to stop behavior, it hides the face of the problem, not the source. Legally it is a hard case too, because where an intermediary's liability for user conduct begins depends on notice, moderation systems and speed of response. On a game for adults, built with the faces of teenage stars, that question stayed the most uncomfortable one of the year.

China's question is older and more structural. In May 2026 PUBG Mobile shut down in China over licensing problems. Tencent then moved players toward Peacekeeper Elite, a game whose gameplay, graphics and content had been substantially altered but which shared many elements with PUBG Mobile. In 2026, ZDNet Korea reported that in its pre-IPO registration filing KRAFTON publicly acknowledged for the first time that it provides technical services for Tencent-operated Peacekeeper Elite and receives service fees for them.

That disclosure revived an old argument. KRAFTON has consistently said the two are separate products. One camp argues Peacekeeper Elite changed enough to satisfy licensing requirements; another asks whether it is a way of keeping PUBG alive in China. The technical service invoice and the franchise ownership are two different documents, but both stand on the same server. I sit in internet cafes in Shanghai and watch this dual identity daily: two names under one regulator's shadow, with the same complaint echoing from Mili — who is telling the truth. My American immigrant eye is not neutral here, and I admit it; an outsider sees the finished product and never the licensing file.

The freshest crisis sits on the esports stage, and this is where six years of casting and live-server observation pay off. At PUBG Asia Stars 2026, two Vietnamese professionals were caught up in it: Himass of Anyone's Legend and TanVuu of The Expendables. On September 23, KRAFTON published its investigation findings and determined that both had used external information, including other players' livestreams, to form judgments and build tactics during matches. In the commission's language this was stream sniping, violating PUBG: Battlegrounds operating policy and the conduct standards for professional players.

The outcome was severe. Both accounts were permanently banned, and both players were barred from all official PUBG Esports events organized or approved by KRAFTON, including PGC, PGS and PNC. The episode began with an accusation from Gen.G's Korean streamer Soopi, who claimed opponents were watching streams. The organizers then removed the two from the rest of the tournament, adjusted points, increased stream delay, cancelled the third match day and split the prize money evenly.

The Vietnamese PUBG community's reaction was immediate and enormous. More than 4.1 million signatures were collected in support of Himass and TanVuu. Big names including Độ Mixi, PewPew, Rambo, DjChip and Ngân Sát Thủ protested or removed PUBG. GAM x The Expendables demanded a clear explanation, while Anyone's Legend took a stand for the players' right to be heard.

My read is this: the central question was never whether sniping happened, and I hold no VODs or admin logs that could settle it. The central question is procedural. In a showmatch, the severity of the sanction, the way evidence is published and the uniformity of the standard must be verified together, or the community never settles, because the argument is not about information, it is about belief. The real question here is not sniping; it is whether the same standard was written for everyone on the same day.

4.1 million signatures are not only an expression of anger, they are proof of organizational capacity. The Vietnamese PUBG ecosystem assembled a petition of that size within days, a feat rare even in international sports communities. I no longer say the ban is the story; the story is the silence around the explanation, and how large that silence is can be measured in signatures. KRAFTON later stated that no further violations were found in other cases. The sentence is administratively correct and strategically weak — to a community that believes the information was published incompletely, 'nobody else' is not comfort, it is the seed of suspicion.

The contrarian angle: where I could be wrong

My first doubt is in the data. The commission's investigation says the real rate was 9%, but without knowing which item tier and which period that rate covers, the exact scale of player loss is hard to state. If that 9% applies only to one skin tier, my loss estimate is larger than reality. My second doubt is in the incentive argument: KRAFTON corrected the information quickly, apologized and paid compensation worth roughly 10.9 billion won, which many publishers do not do. If a regulator credits that cooperation and keeps the penalty light, that is not unreasonable in principle — and my 'cheap defense' thesis becomes somewhat unfair.

My third doubt sits in the Vietnamese case, and here I should name my vantage point. I am not an expert in Korean consumer law, and I am not inside Vietnamese esports administration. I write from Shanghai, limited to three things: livestreams available to watch, published community reaction and KRAFTON's official documents. Years spent casting mobile esports in South Asia taught me that when an outsider decides what justice means for a Vietnamese or Chinese community, they usually get it wrong. On one point I have no doubt: the same pattern ran through all three crises — a loud announcement, then a correction, then compensation, and finally an open question about procedural transparency. KRAFTON did not fail the test; it changed the question paper, and the real accounting starts the day the questions stop changing.

Looking forward

The next major PUBG esports dispute will, I expect, be about the practice of publishing admin evidence rather than about item odds. The organizer who first attaches video timestamps and admin logs to a decision will take the first step toward community trust; the one who publishes only a list of sanctions is writing the next petition against themselves.

And for the player, the question stays simple: the next time someone prints 100%, will you believe it or will you calculate it? The answer depends on who kept running through these three crises, and how.