Nearly 300,000 League of Legends and VALORANT Accounts Locked — The Most Worrying Part Is Not the Count

Nearly 300,000 League of Legends and VALORANT Accounts Locked — The Most Worrying Part Is Not the Count

The moment an account vanished

At 11:40 p.m. on October 14, 2026, I was sitting in front of three monitors in a small apartment in Mapo-gu, Seoul, doing what I do every evening: scanning the MMR distribution table of the Korean server. An account in Diamond I, with a 71% win rate across 212 games, sat at number 4,318 in the ladder. By 11:40 that same night, it was gone. No notice, no trace — just a blank space in the data file I save every thirty minutes.

Three days later, on a community forum, the owner of that account opened a thread: "My account got locked for no reason, I did nothing wrong." I read all 412 comments. The community split cleanly in two: one side believed the system had erred, the other left only two words — "paid boosting." I did not join the argument. I saved the screenshot, noted the timestamp, and waited, because after thirteen years observing this industry I know that a large-scale purge always leaves traces before it is announced.

A week later, those traces became a global headline: Riot Games announced it had actioned nearly 300,000 League of Legends and VALORANT accounts for behaviour connected to ranked play.

I read that announcement twice. The first time as an analyst. The second time as someone who was once called a traitor to a historic victory simply for publishing an honest xG figure.

Vanguard travels from VALORANT into League of Legends

To understand this enforcement wave, you have to start with the thing the headline never mentions: Vanguard.

Vanguard is a kernel-level anti-cheat client. It launched alongside VALORANT, and according to Riot Games' announcement it was integrated directly into the League of Legends client by September 2026. That is the pivot most coverage skipped. Riot did not merely port a tool from one title to another; they moved Vanguard from the role of detecting cheat software to the role of governing behaviour inside the ranked system.

That difference is bigger than it looks. Detecting cheat software is a pure engineering problem: either the malicious code is there or it is not. Governing ranked behaviour is an interpretive problem: who is playing, playing for whom, playing alongside whom, and with what motive. One can be solved by an algorithm. The other requires a judicial system.

For years in the data room I have reminded myself of one thing: when an organisation is simultaneously the rule-maker, the enforcement body, and the sole publisher of data about itself, that data must be read with a different posture. Not default suspicion — disciplined caution.

The second important context: the ranked ladder is not casual entertainment. It is the de facto selection system for the entire amateur-to-professional pipeline. Academies and tier-two organisations across Korea, China, Europe and North America all use solo/duo standing as a screening filter. When that filter is polluted, what breaks is not the ordinary player's experience but the talent-identification signal.

That is why I followed this story with the same focus I usually reserve for a transfer window.

How much is three hundred thousand, really

Before you trust a number, ask where it was born.

The figure of 300,000 sounds enormous. But the same report supplies a corrective ratio: roughly 0.2% of the combined monthly player base of the two titles. That ratio is derived from two estimates — around 120 million monthly League of Legends players and around 20 million VALORANT players, roughly 140 million combined.

There are three provenance problems I am obliged to raise, because this is my job.

First, neither the 120 million nor the 20 million figure has a stated source. They appear as "estimates show" and "said to be." In any report I ever filed to the data desk in Seoul, an unsourced number was returned immediately, however plausible. Not because it was wrong, but because it could not be verified.

Second, the time window behind the 300,000 figure is undefined. The Vanguard integration into the League client dates to September 2026. If enforcement began then, the 300,000 represents roughly one quarter of cumulative action, not a year. Annualised, the intensity is materially higher than the first impression suggests. Conversely, if the figure folds in earlier VALORANT enforcement, it is diluted across two titles whose player bases differ sixfold. Two readings, two opposite conclusions — and the report gives us no tool to tell them apart.

Third, and this is the point I consider most important: the denominator is almost certainly wrong.

I will give that its own section, because I once lost three nights of sleep over a denominator.

The hidden denominator: the China problem

League of Legends and VALORANT in mainland China operate inside Tencent's ecosystem, with localised anti-cheat and account-verification infrastructure distinct from the global Vanguard rollout. That is a structural fact, not speculation.

If the 300,000 figure reflects global servers excluding China, then dividing it by a 140 million denominator misstates the 0.2% ratio in a direction favourable to the party publishing it — because a very large share of monthly League of Legends players sit inside the China ecosystem, and that population may fall outside this enforcement wave.

I am not claiming the 300,000 excludes China. I am saying the report does not say, and that silence carries analytical value. In valuation work we call this a denominator error, and it is more dangerous than a numerator error, because it never shows up in the headline.

When I opened my regular online seminar — the Discord channel I built during the empty-stadium season of 2026 — to ask members how they read the figure, 217 people responded. Sixty-eight per cent said they remembered only the 300,000 and never noticed the 0.2%. Only nine per cent recognised that the ratio depends on an unsourced denominator.

Data does not shout, it whispers — and I have learned to lean in and listen. But most audiences do not lean in. They only hear the headline scream.

"Hitchhiker" — the new doctrine and its price

This is the part that made me reread the original text three times, and the part I expect to be most contested over the next six months.

The report identifies a category with a striking name: "hitchhikers." These are players using their own accounts, playing every game with their own hands, but queuing alongside an account that is being boosted. Under the new policy they may lose ranked points earned in affected games, even though they broke no software rule themselves.

I understand the logic. Leave the companion untouched and the boosting market simply needs one "clean" account as a chaperone, neutralising every measure within weeks. But there is a gap between understanding the logic and accepting it, and that gap is liability by association.

In thirteen years writing about sport, I have watched exactly one system expand liability to third parties, and it did not end well. The Seoul night of 2026 taught me that truth can be lonely, but it is never wrong. That lesson applies to the accused and to the accuser alike.

The report contains no information on false-positive rates, on an appeals process, or on the evidentiary standard used to classify someone as a hitchhiker. For an enforcement wave touching nearly 300,000 accounts, that absence is a substantial transparency gap.

Consider a concrete case. An ordinary player queues every night with a close friend. That friend, for reasons of his own, paid a boosting service to climb from Platinum to Diamond. The first player does not know. After the enforcement wave, he loses an entire season of ranked points, with no channel to explain himself.

That is the scenario I want answered with data, not with a line in a terms-of-service document.

Smurfing is not automatically cheating

There is a detail in the report that I believe gets skimmed most often and carries the most information about Riot's enforcement philosophy.

Riot Games representative Phillip "mirageofpenguins" Koskinas addressed the smurfing question. The gist: smurfing is not automatically treated as cheating. Riot enumerates eight legitimate uses of secondary accounts, including players wanting to protect their highest achievement on their main account.

Eight cases. Not one, not two. Eight.

That detail transforms how I read the whole enforcement action. Riot is not building policy on account count. They are building policy on behaviour and intent. You can hold five accounts and remain legitimate, if your purpose sits inside those eight cases. You can hold two accounts and be actioned, if your purpose is raising someone else's rank for money.

This is a deliberately soft boundary. And soft boundaries have a cost: consistency. I have tracked many sports-governance systems, and the pattern is always the same — intent-based rules are harder to apply evenly than measurement-based rules. When the same behaviour can be read as "protecting my personal achievement" or "evading the ranked system," the judge needs a great deal of context, and context is never distributed fairly across regions.

Based on my experience tracking matches and governance structures, this will be the next battleground — not in the West, but on servers with high secondary-account density and a strong climbing culture, Southeast Asia being a clear example.

The unheralded gift: LP protection

Among the flood of account-ban news, one detail was overlooked that I believe will change the daily experience of tens of millions of players more than any ban.

When the system detects that a match contained a cheater or a leaver, affected players do not lose ranked points for that game.

That sounds small. Look at it through the eyes of a valuation analyst and it is enormous.

Previously, every ranked game was a noisy expected-value calculation. You spend thirty minutes and a slice of mental energy, and you receive a point swing that depends on four teammates you do not control. When one of those four is a cheater or a leaver, your investment is written off with no refund mechanism.

LP protection turns that game into an insured investment. Mathematically, it compresses the variance of a match sequence. Over the long run, it makes ranked points a marginally more accurate skill signal — which is precisely what the entire scouting pipeline needs.

I want to stress this because it is underreported yet structurally valuable: a clean ranked system is not the one that bans the most, but the one with the smallest variance between true skill and displayed rank.

TPM 2.0 and the two-tier citizenship model

The most consequential part of this whole story sits in the future tense.

The report mentions Riot Games' plans: multi-factor authentication, TPM 2.0, hardware authentication and — the most striking detail — verification requirements that may be applied differently depending on a player's rank.

TPM 2.0 is a hardware security standard enabling device-level identity attestation. In plain language: your account becomes bound to a specific machine. When that happens, the cost of creating a "one-time" account spikes, and Riot's stated goal is to make such accounts harder to sustain.

As anti-cheat efficacy, I believe the plan will work. As governance, it creates two problems the report never touches.

The first is privacy. Binding accounts to hardware intersects directly with personal-data regulation in several jurisdictions. This is not a theoretical worry; it is the kind of issue that has paused biometric verification systems in multiple markets.

The second is two-tier citizenship. If higher-ranked players must verify more rigorously than lower-ranked ones, that is logically sound — fraud risk concentrates at the top of the ladder, where boosting money flows most and where scouts are looking. But in principle it creates two classes of player inside one game.

And there is a third, more practical problem: shared devices. In many regions, a significant share of League of Legends players access the game through internet cafés, shared family computers, or older hardware without TPM 2.0. If hardware attestation becomes a requirement at higher ranks, a group of legitimate players is pushed out of the upper ladder not by skill, but by material circumstance.

That is a new form of inequality, and it appears nowhere in the report.

Gray-market economics: repricing, not disappearance

Now the counterintuitive part.

The community's first reaction was: "Finally, cheating is getting smashed." Understandable — but it ignores a basic law of gray markets that I have observed long enough to trust.

Boosting exists because there is demand and supply. Demand comes from players who want a rank their skill cannot produce — for prestige, for seasonal rewards, for ego, for a nice screenshot. Supply comes from highly skilled players who need income, and at the bottom of the esports labour pyramid, income is a serious problem.

Enforcement does not erase demand. It does not erase supply. It raises risk, and risk has a price.

If you raise the cost of an activity without touching demand, the market responds in one of three ways: higher prices, relocation, or migration to another title. I expect all three. Boosting prices in League of Legends and VALORANT will rise. Some activity will move to titles with softer verification, where risk cost is lower. And some will move to closed channels, where monitoring becomes harder for Riot and for data analysts like me.

This does not make the enforcement meaningless. The opposite. It means the problem is industrial, not title-specific. When one publisher raises the bar, gray activity does not vanish — it migrates.

The transfer market is a magic trick: look closely and you see the strings. The boosting market is the same, except the strings are longer and run back to economic problems no ban can solve.

The fan's view

Since 2026, after the article about Korea beating Germany that got me branded a traitor, I have always closed my analyses by listening to fans before concluding. This time I asked them, and the result forced me to revisit some of my own assumptions.

Across the 217 responses I collected, the most notable thing was not the split for or against — it was the silence about smurfing policy. Very few respondents mentioned the eight legitimate cases. Most asked one question: "So if my friend queues with a boosted account, do I get banned?"

They do not care about the 300,000. They care whether the risk can land on them. That is what I learned in 2026: data must be framed with empathy, or it is just a talking spreadsheet.

I do not stop you from placing a bet — I only want you to understand what you are betting on. I still use that line about match odds. This time it applies to something else: when you queue ranked with friends, you are also betting on a governance system you cannot fully see.

What I am watching in the next cycle

I will not close with a summary, because summaries serve the writer, not the reader. I will tell you what I am waiting for.

First, whether Riot Games publishes enforcement data on a recurring cycle. If it does, esports gains a transparency precedent other publishers must weigh. If it does not, we are left with one-off numbers, no trend line — and numbers without a trend line cannot be used to judge anything.

Second, the post-enforcement rank distribution. If locked accounts really concentrate at the top of the ladder, as rank-tiered verification implies, we should see a temporary contraction in the high-elo population and some distortion in MMR calibration benchmarks. That would be the first independent evidence of effectiveness not sourced from the enforcer itself.

Third, price signals in the gray market. But that is observation for understanding, never for participation.

Finally, I am waiting to see who — a regular player, a pro, a journalist — steps forward to ask about the appeals process. In any judicial system, what protects the innocent is not severity. It is transparency.

With no audience in the stadium, I hear the match breathing. The same applies here: the most important sound is not 300,000 lock orders. It is the silence around one question — if this system is wrong, who fixes it?