296,416 Ghosts Beneath the Ladder: How Riot Games Builds the Anti-Boost Shield

Mouse Clicks in a PC Bang at 3 A.M.

In Busan, I grew up between two sounds: the whistle of the Line 2 subway and the click of mechanical keyboards falling like rain inside the seaside PC bangs. When I was seventeen, on a winter night, I sat in the innermost corner of an internet cafe near Seomyeon Station, watching a boy of about fifteen play VALORANT on an account that was not his. He typed so fast I thought his hands would leave the desk. On screen, a Reyna dashed into the bombsite and took down three players in four seconds. It was breathtaking.

Then the clock struck three. The boy shut down the machine, bowed to the shop owner, walked out under the yellow streetlights. On the console, I saw that the account he had just logged into still carried a small line of text: Last logged in from a different device. He was not the account's owner. He was a hired climber — a booster, a ghost-painter for a ladder that did not belong to him.

I tell that night not to judge. I tell it because it helps me understand what Riot Games is trying to do, and what they may never be able to do.

A Shield Called Anti-Boost

According to Riot Games' official communications, the Anti-Boost system — the shield against boosting behavior — has actioned 296,416 accounts exhibiting rank manipulation behavior across both VALORANT and League of Legends. The figure was released in a bulletin issued by Riot's communications team over a window described as "from late last year to now," without a fixed absolute date for each statistical sub-period.

This is the largest number Riot has ever published in a single release on this topic. But like every number coming from a publisher itself, it should be read with two eyes: one trusting the controller's intent, one doubting the possibility of independent verification. I am not writing this article to celebrate or to dismiss. I am writing because behind that figure lies a complex governance system, a four-tier penalty ladder, a joint-liability model, and — most importantly — gaps that neither the community nor Riot itself has yet named.

This article is for anyone who has climbed ranks in silence, suspected an opponent was someone else, and wondered whether anyone behind the ladder was counting them.

Context: The Ladder as a Living Organism

To understand Anti-Boost, one must first understand what it is trying to protect: the ranked ladder. This is not a static list. It is a living ecosystem, where every match shifts a player's position, where skill is priced in points, and where personal reputation is bound to a public hierarchy that cannot be faked — at least not technically.

Boosting breaks that very assumption. It lets a high-skill player log into someone else's account, climb the ladder for them, and return the account at a position its owner never reached by their own hand. Purely technically speaking, it is a trade in moments — you do not earn it through effort, you pay someone to earn it for you.

Throughout this article, I will use English terms for precision: boosting, booster, smurf, deranking, and rank manipulation — the umbrella term for buying, selling, transferring accounts, intentional losing, and using another person's account to climb.

Vietnamese VALORANT players call boosting by a shorter name: cày thuê. But whatever you call it, the problem is the same: it breaks the promise every ranking system makes — that the better player sits above the weaker one, and that your position is the truth about you.

How Riot Defines Violating Behavior

In its official material, Riot Games splits violating behavior into a clear taxonomy. This is an important point, because it shows that this publisher does not conflate every alt account with every cheating behavior.

First, there is directly detected rank manipulation — an account being operated by someone other than its owner, for the purpose of climbing the ladder for the true owner.

Second, there is account buying, selling, and transferring. This is a commercialization of rank: you pay to obtain an account at a desired rank, or you sell your own account to someone else.

Third, there is intentional deranking. According to Riot, this is a severe behavior because it not only ruins the experience of four teammates, but also facilitates subsequent forms of manipulation — for example, dropping rank to create a smurf at a low tier, or entering low-competition matches for profit from boosting services.

Fourth, there is using a secondary account to gain competitive advantage — often called smurf-assisted climbing.

Notably, Riot does not ban alt accounts in general. This is a very carefully drawn boundary, and it carries significant meaning across the entire governance system. According to official explanation, a player creating and self-operating their own alt account is normal activity — what Anti-Boost targets is not the existence of alt accounts, but the intent to manipulate rank. This is an intent-oriented standard, and it differs substantially from bright-line bans.

When I first read this section, I remembered an old story from Busan. Summer 2026, I had a friend named Min-jae with three League of Legends accounts. Min-jae never boosted, but he used an alt to play with his cousin in Silver. No one punished him. Riot understands that. But if Min-jae handed his account to a Challenger to climb for him to impress his girlfriend, that would be different. The difference lies in intent, not in the act of logging in.

The Four-Tier Penalty Ladder: From Rollback to Permanent Ban

The most intriguing part of Riot's material is how they designed the penalty ladder. It is not a single punishment for all violations, but a structured escalation system.

Tier one covers detected manipulation cases. Penalties include: cancellation of ranked points and rewards gained from cheating, returning the account to its original rank, and temporary suspension. This is a rollback-type penalty — it tries to restore the state that existed before the violation.

Tier two covers repeat offenses. Here, ban duration increases on an escalating scale. Riot does not publish the exact escalation formula, but the principle is clear: the more times you violate, the heavier the penalty.

Tier three covers account buying/selling or intentional deranking. According to the material, these behaviors may lead to permanent bans. This is a notable step up, placing these two behaviors in the most serious category of the taxonomy.

Tier four — and this is the most controversial part — extends penalties to related parties. Specifically, the booster's main account may also be actioned, and according to the material, players who frequently party with a booster may also fall under enforcement.

In English, this is called joint liability. In Vietnamese legal language, the closest concepts are đồng phạm or liên đới trách nhiệm, but in the online-game context, it takes on a different meaning: when you frequently play with someone, the system may treat you as part of the same violating pattern.

I have a friend who plays VALORANT named Huy, in Hanoi. He once duoed with a close friend for a month, climbing from Gold to Platinum. That friend never boosted, but at one point Huy received a warning message from Riot about "signs related to rank manipulation" and his account was suspended for three days. Huy panicked and called me at midnight. After checking, it turned out that Huy's duo partner had an alt account under investigation. Huy did nothing wrong, but Huy was swept into the vortex of joint liability.

This is not an isolated story. It is the direct consequence of how Riot designed its policy.

The Detection Mechanism: When a Machine Reads Behavior Instead of Evidence

One of the points I always want to analyze deeply is Anti-Boost's detection mechanism. Riot's material does not disclose algorithm details, which is reasonable — disclosing them would teach boosters how to evade. But from how they describe it, some inferences can be drawn.

First, the system operates at the account and behavior layer, not the game-balance layer. This means Anti-Boost is independent of champion or map balance patches. It does not care whether you are playing Reyna or Jett; it cares whether you are playing your own account.

Second, the system is reactive-with-rollback, not fully preventive. This means there is always a lag between when manipulation occurs and when it is actioned. A booster may climb dozens of matches before being caught. During that window, other players — teammates and opponents alike — have to endure the consequences.

Third, Riot says it is expanding the system in two directions: (1) scaling enforcement, and (2) developing the ability to detect boosting signs at the match level. The second direction is particularly important, because it shifts from reading accounts to reading play patterns — for example, an account suddenly playing with the style of a Challenger after months of only Silver play.

This reminds me of how anti-fraud systems in traditional sports operate. In football, VAR cannot detect a match-fixer through one play; it needs behavioral chain analysis across many matches. In tennis, match-fixing authorities monitor abnormal betting patterns rather than just outcomes. Anti-Boost is walking the same road: it does not try to catch in the act, it tries to recognize traces.

And when you build a system based on traces, you must accept one truth: you can never achieve perfect accuracy.

A Number That Does Not Reveal a Trend

In this article, I want to dedicate a substantial section to reading the figure of 296,416 accounts through a statistical lens. Because this is the number the media keeps repeating as proof that "Riot is tightening the screws."

But the data does not say that.

296,416 is a cumulative total, not a trend indicator. To know whether Riot is tightening the screws, we need at least two data points: a baseline and a comparison point. The article provides no baseline. It provides a single figure, with no breakdown by game, no breakdown by region, no breakdown by time.

Even more, this figure pools VALORANT and League of Legends together. This is a serious analytical problem, because these two titles have very different booster economies. VALORANT is a tactical shooter where individual skill is strikingly visible and where a high-skill player climbing on someone's behalf can create enormous match impact. League of Legends is a MOBA where results depend more on team coordination, and where the prestige of a high rank can carry greater social value. Boosting demand differs between the two titles. Rank-inflation pressure also differs. Yet Riot pools them into one number.

When you pool two different things into one number, you do not make the information stronger. You make it more ambiguous.

Moreover, this number is self-reported by Riot. No independent auditor confirms it. In the world of platform governance systems, this is not unusual — major tech companies routinely self-report their enforcement statistics. But as a journalist, I must say it clearly: this is a self-reported figure, not independently verified.

The Most Dangerous Blind Spot: The Frequent-Teammate Clause

Now I want to dedicate this section to what I consider the most dangerous blind spot in the entire Anti-Boost design: the clause handling players who frequently party with a booster.

In theory, this clause has logic. If a professional booster plays frequently with a group of friends, that group may be part of the same violating ecosystem. Applying joint liability could break that ecosystem more effectively than banning one isolated account.

But in practice, this clause creates an enormous risk zone.

Imagine an ordinary player. She plays VALORANT every evening, duoing with a close friend from university. That friend has a main account played legitimately, but occasionally uses another account to play with a younger cousin. She knows nothing about that account. She only knows her friend. Then one day, her account is suspended for "signs related to rank manipulation."

Can she appeal? Riot's material does not describe an appeal mechanism specifically for joint-liability cases. No tolerance threshold is published. No clearance mechanism is stated.

This is the kind of risk I call structural false-positive risk. It is not a system error in a specific case — it is the consequence of how the system is designed. When you extend liability to related parties, you accept that some innocent people will be affected. The question is: how many innocent people is acceptable?

Riot has not answered that question.

Who Holds Accountability

Another important governance aspect I want to raise: across the entire Anti-Boost system, Riot controls both detection and adjudication. No independent third-party appeals body is mentioned.

Compare with traditional sports. In football, when a player is sent off, the club can appeal to the league organizer. If the organizer does not resolve it, they can appeal to a higher governing body — e.g., FIFA or the Court of Arbitration for Sport. There are multiple independent review layers.

In esports, and especially in online skill-based ranking systems, this architecture does not exist. Riot is simultaneously legislator, police, and judge. This is not necessarily bad — it is how the game industry has operated. But it means transparency depends entirely on the publisher's goodwill.

I am not saying Riot lacks goodwill. I am saying a governance system resting only on goodwill cannot be considered absolutely reliable.

The Gray-Market Economy: When Rank Becomes a Commodity

To understand why boosting exists, we need to look at the gray-market economy behind it.

Rank in VALORANT and League of Legends carries a certain social value. It is a form of digital prestige, a skill signal that can be displayed before friends, before a community, and sometimes before employers. A Challenger or Radiant account can be used to build personal branding, to attract streaming views, or simply to feel more confident in matches.

When something has social value, it has economic value. When it has economic value, it spawns a market. And when there is a market, there are service providers.

Boosters provide rank-climbing services. Buyers pay to obtain a rank they cannot reach themselves. In between are intermediary platforms, closed communities, and payment systems sometimes sophisticated enough to avoid detection.

Anti-Boost impacts this market in two ways. First, it raises the expected cost of getting caught for both buyer and seller. If you are a buyer, you do not just lose money on the service; you risk permanent account loss. If you are a seller, you risk losing your main account. Second, it directly attacks the account-market supply side by permanently banning buying and selling. This is a supply-side measure more than a demand-side one.

But Riot's material provides no data on market size, recidivism rate, or the actual economic impact of enforcement. We know 296,416 accounts were actioned. We do not know whether that represents a small or large share of all accounts engaged in the gray market.

And when the denominator is unknown, the numerator cannot be evaluated.

The Asymmetry Between Detection and Evasion

A structural feature of every online anti-fraud fight is the asymmetry between the speed of developing detection methods and the speed of developing evasion methods.

Riot invests in Anti-Boost. It scales the system. It develops match-level detection. But boosters are also learning. They change communication, team-up patterns, payment methods. They move from easily monitored channels to harder-to-monitor ones, from simple to complex models.

This means part of Anti-Boost's effectiveness will always erode over time. When Riot publishes a large number, it says: "We caught many." But the more important question is: "How many were not caught?" Riot does not answer that.

In sports, this is called the iceberg problem. You only see the tip. The submerged part — undetected behavior — may be far larger. Anti-Boost is trying to melt the iceberg, but seawater is still freezing from below.

What Anti-Boost Is Doing Right

I do not want this article to sound like wholesale criticism. There are things Riot is doing well, and they deserve recognition.

First, not banning alt accounts in general is a correct system-design decision. Other platforms have tried banning all alts, usually to chaotic results for legitimate players with genuine multi-account needs. Riot distinguishes between the existence of alt accounts and the intent to manipulate rank. This is a hard line to draw, but necessary.

Second, distinguishing penalties by severity is sensible. Not every violation deserves a permanent ban. By placing permanent bans at the top of the ladder, Riot sends a clear message: the heaviest penalty is reserved for the most commercially driven and system-destroying behaviors.

Third, extending detection to the match level shows Riot is thinking long-term. It is not just trying to handle individual accounts, but trying to understand behavior patterns at the system level. This is a correct step.

Fourth, publishing the 296,416 figure — despite lacking context — is still a signal of transparency. Many publishers publish nothing. Riot chose to publish. That is a noteworthy event.

When I think about these points, I remember a Sunday afternoon in Busan, sitting watching a VALORANT stream, reading a comment from a player: "At least they are trying." That simple comment contains a complex truth: we do not need a perfect system, we need a system that is trying to be better.

The Contrarian Angle: When Anti-Cheat Breeds Distrust

Now to the part I want to call the contrarian angle.

Every anti-cheat system has an inherent paradox: the stronger it is, the more it breeds suspicion among the honest.

Imagine a very sensitive Anti-Boost. It detects every anomaly. It actions every suspicious account. This sounds good — until you realize that every sensitive system has a certain false-positive rate. And when that rate is applied to tens of millions of players, the number of innocents affected can reach thousands.

In Vietnam, the VALORANT community sometimes shares stories of accounts suspended without clear reason. These stories may be true, false, or misunderstanding. But their existence signals that the community lacks a transparent information channel about how the system works.

This paradox is not unique to Riot. It exists in every platform governance system. But when a publisher chooses to publish enforcement figures, they should also publish error rates. A number like "we actioned 296,416 accounts and the appeal success rate is X%" would carry far higher informational value than a bare number.

The Community's Silence: A Gap in the Picture

One notable thing in Riot's material is the total absence of community feedback. No voice from actioned players. No independent organization speaking up. No debate mentioned.

This is an information gap. It does not mean no feedback exists — it just means the material we are analyzing does not include it. But the very absence signals the material's origin: it comes from the publisher, not from an independent journalistic platform.

In sports journalism, we usually have two types of information: information from decision-makers and information from those affected. A balanced article needs both. The Anti-Boost piece I am analyzing has only one.

This reminds me of an evening in Busan, having coffee with a former esports pro. He told me something I still remember: "The problem with esports is not a lack of rules. The problem is that rules are often written by people who never have to endure them."

I do not fully agree with him. But I understand what he means.

What Comes Next

Riot says it will continue scaling enforcement and developing detection. This means we will see more enforcement waves in the future, and possibly more violation categories added to the taxonomy.

In the short term, this is good for ladder health. In the medium term, it raises the question of how Riot will handle new challenges: more sophisticated evasion tactics, cross-border boosting markets, and community pressure for greater transparency.

In the long term, the biggest question is: can a ranking system maintain integrity in an economy where rank has monetary value?

I do not have the answer. I only know the answer will not come from a publisher's bulletin. It will come from millions of small decisions by millions of players, each choosing where they want to stand on the ladder — and how.

Lessons From a Boy in a Busan PC Bang

Back to that winter night in the PC bang near Seomyeon Station. The fifteen-year-old boy I saw, who played VALORANT on an account that was not his, has likely grown up by now. I do not know whether he kept boosting. I do not know whether Anti-Boost actioned him. I do not even know his name.

But I know one thing: he played very well. His skill was real. What was not real was the position he was helping someone else reach.

When I think about Anti-Boost, I think about that boy more than the number 296,416. The number is just a full stop. The boy is a question mark. And question marks are what keep a journalist awake at 3 a.m.

What a System Cannot Do

There is a structural limit that Anti-Boost, however improved, cannot overcome: it cannot create a sense of responsibility. It can only create a fear of detection.

Fear of detection can reduce the number of violators in the short term, but it does not change why people violate. It raises the cost of violating behavior, not the demand for it. It pushes the gray market deeper, better hidden, more professional.

This is not Riot's fault. It is the nature of any deterrence-based governance system. We see it in traditional sports with doping, in finance with tax fraud, in games with cheating. There is always a race between anti-fraud and fraud, and always a gap between rules and reality.

This does not mean we should give up the fight. It means we should understand that a good system is not a perfect one, but one that knows its own limits.

Does Riot know its own limits? They know detection needs improvement — clearly stated in the material when they discuss match-level detection. But do they know transparency is also part of the limits? That is not yet clear.

When the Ladder Becomes a Sanctuary

Over six years writing about esports, I have learned that the ladder is not just numbers. It is a small sanctuary each player builds for themselves. It is where they bet on their own skill, measure their own progress, find friends and rivals, and sometimes find themselves.

When a booster invades that sanctuary, they do not just ruin a match. They ruin a belief. The belief that if I try, I will be rewarded fairly. The belief that if I lose, it is because I am not good enough, not because someone is playing for someone else. The belief that the ladder is a truth, not a transaction.

That is why I write this article. Not to praise Riot, not to criticize them. But to remind that behind every governance number are people trying to believe in something.

Those Who Are Never Named

I always write about those not named in the victory table. In the Anti-Boost story, who are they?

They are players forced to play with a booster unknowingly. They are players defeated by a booster, wondering why they were so bad. They are duo friends caught in joint liability for playing with someone they trusted. They are new players climbing in silence, unrecognized because the system is busy dealing with manipulators.

And somewhere among them, there may be a boy in Busan, playing very well, but on an account not his own, because that was the only way he found to make his skill valuable.

I am not defending his behavior. But I understand why it exists. And I believe a good governance system must understand that before it tries to change it.

What I Want to See in the Next Shield

If I could send one proposal to Riot Games, I would propose three things.

First, publish enforcement data separately per game. VALORANT and League of Legends have different booster economies, and pooling them into one number loses analytical capability. A split number would carry far higher informational value.

Second, publish a tolerance threshold for the joint-liability clause. No need to reveal algorithm details, but players need to know there is a clear threshold — for example, "we only action accounts that partied with a booster over X matches in Y days." This would help honest players understand they are not in a high-risk zone.

Third, publish the appeal success rate. This is an important indicator of system quality. If the appeal success rate is high, the system has an accuracy problem. If it is low, the system is working well. But whether high or low, players have a right to know.

These three proposals do not require Riot to reveal technological secrets. They only require Riot to treat the community as a partner in the fight against fraud, not merely as an object of management.

A Question Left Behind

I close this article with a question I cannot answer.

If you are a player climbing the ladder in silence, and you know your duo partner sometimes plays on another account, would you report them?

If you know a friend is boosting to pay tuition, would you report them?

If you know you could be actioned for joint liability, would you stay silent?

These are not easy questions. And I believe any anti-fraud system must face them, however well designed.

The ladder will keep living. The rungs will keep being climbed. The ghosts will keep appearing under account names that do not belong to their players. And somewhere, in a PC bang in Busan or a dorm room in Hanoi, a young person will be trying to find their true position — one way or another.

What I hope for is not a perfect system. It is a system honest about its own imperfection.