296,416 Accounts and the Architecture Behind Them: How Riot Games Rewrites the Rules of the Boosting Market

On a Saturday night, in an internet café in Binh Tan District, a seventeen-year-old boy wired two and a half million dong to a stranger's account. The deal was brutally simple: push his account from Silver to Platinum within seven days. At the same moment, on the other side of the hemisphere, Riot Games' servers were logging every anomalous play made on that account.

Three months later, the account vanished. Not because the boy was hacked, but because Riot concluded that the rank points he held were the product of an illegal transaction and had to be wiped clean.

That is how Anti-Boost operates. Not a blanket ban on every kind of alt account, but a mechanism aimed squarely at the intent to manipulate rank. And when you look at the cumulative figure of 296,416 accounts penalized across both VALORANT and League of Legends, what matters isn't the scale — it's the architecture behind it.

Four tiers of punishment and an expanded liability model

Riot did not build Anti-Boost as a thunderbolt from the sky. They built it as a ladder.

Tier one, for first-time detection of manipulation: rank points and rewards earned through cheating are cancelled, the account is returned to its original ranking, and a temporary suspension is applied. This is a retroactive penalty — it does not prevent the behavior from happening, it erases the traces of the behavior after it has already happened.

Tier two, for repeat offenses: ban duration escalates. The more it recurs, the harsher the sentence.

Tier three, for two more serious categories of violation: buying, selling, or transferring accounts and intentional deranking. Here, Riot is willing to hand down a permanent ban. This is notable in terms of the economics of punishment: Riot classifies these two behaviors as heavier than plain boosting, because they are links in an organized black market rather than the isolated act of an individual.

Tier four is the most contentious part. Riot does not only act on the manipulated account. They extend liability to the booster's main account, and to teammates who frequently queue alongside that person.

I spent a fair amount of time digesting that last clause. On a system-design level, it makes sense: if a duo has queued together for hundreds of matches and one of them is a booster, the assumption that the other knows nothing is a weak assumption. But operationally, it opens a gray zone. No tolerance threshold is published. No independent appeal mechanism is described. Riot is simultaneously the investigator, the judge, and the executioner.

The blind spot of an intent-based standard

Before going deeper, a clear boundary must be drawn, one that Riot itself has drawn. Alt accounts created and operated by the player themselves are normal activity. Anti-Boost does not target the existence of alt accounts; it targets the intent to manipulate rank behind their use.

This is a narrow standard by design, and that very narrowness is what creates the problem. A bright-line rule — for instance, banning all alt accounts — would be easy to apply, easy to make transparent, easy to explain to the community. But it would also kill legitimate activity. An intent-based standard protects legitimate players, but in exchange, it depends on behavioral signals that both players and observers struggle to verify.

The consequence: when the system wrongly punishes someone, that person has almost no way to prove their innocence, because the evidence sits in the hands of the accusing party itself. No independent appeal body is mentioned anywhere in the information Riot has published.

I am not saying this to accuse Riot of wrongdoing. I am saying it to point out that a system that accumulates the power to investigate, adjudicate, and enforce into a single entity always carries a structural blind spot. That blind spot does not disappear when the entity has good intentions. It merely waits for a loud enough case to expose itself.

When intent is scored by algorithm

There is a psychological effect that behavioral researchers call the asymmetry between perceived risk and the actual probability of risk. Boosters rarely assess their odds of getting caught accurately. They look at the matches they have successfully boosted, not at the accounts that have vanished.

The figure 296,416 has a peculiar quality. To a news reader, it is a milestone. To a player weighing a paid boost, it is a number too large to picture, and therefore easy to process as a meaningless number. Numbers can weep, if we are willing to listen — but most people do not listen, because they do not believe the number includes them.

What is notable is that Riot publishes this number itself. No independent auditor verifies it. I have no reason to believe Riot is fabricating it, but it must be said plainly: this is self-reported data, not cross-verified data. A milestone with no comparative baseline — 296,416 over what period, against what previous cycle — cannot be used to prove a trend. It proves a total, not a direction.

And this is where I have to be honest about my own profession.

I was wrong, so I can see the map

In 2026, while writing for a sports outlet, I covered a wave of account bans in a tone dripping with prophecy. I called it a comprehensive purge. I wrote as if everything had been settled. A year later, looking back at the data, I realized I had been wrong in exactly one place: I turned a total into a trend, and turned a publisher's statement into a verified fact.

That lesson shaped how I read communiqués like this to this day. When Riot announces it will scale up Anti-Boost, including a new detection layer based on match-level signs of boosting, that means the current methods still have holes. You do not build an additional detection layer for a problem that has been solved conclusively. You build it because the problem is still running.

There is an asymmetry worth naming. Riot expands its detection capacity. Boosting crews adapt. This race has no finish line, and the claim of an ever-tightening crackdown is an inference, not a datum. That is why I always hunt for a number that contradicts the very argument I am building: if behavior-signal methods are not yet mature, then the false-positive risk is a real risk, not a theoretical concern.

Why this matters more than it appears

There is a story seldom told. Ranked ladders in competitive titles are not merely places where players unwind. They are the input-data layer for the talent-scouting pipeline. Academy teams, academies, scouts — they look at the ladder to find traces of individuals with potential. If the ladder is inflated by boosted accounts, the quality of the signal scouts receive is diluted.

This is why I argue that a clean ladder is not a small operational matter, but infrastructure for the entire scouting ecosystem. You cannot build a talent-detection system on a foundation where rank can be bought with money. Esports is teaching football how to speak the language of a new generation — and the first lesson in that language is: data is only trustworthy when it cannot be bought.

More broadly, the way Riot operates Anti-Boost is a model worth referencing by other publishers. They own the entire chain from detection to punishment. They publish a clear violation taxonomy, an escalating penalty ladder, and a liability model that extends to related parties. This is a statement of sovereignty: whoever operates the ecosystem sets the law. The strongest is not the fastest runner, but the one who reads the wind of the market.

But sovereignty always comes with corresponding responsibility. If Riot wants the community to trust Anti-Boost's fairness, publishing a cumulative figure is not enough. It needs a comparative baseline. It needs a tolerance threshold for the joint-teammate clause. It needs an appeal mechanism where players can bring their evidence to a forum not organized by the accusing party itself.

What remains after the number

When all is said, what I keep is not the figure 296,416. It is a much smaller detail: a player who deleted his own alt account after reading the new violation taxonomy, afraid that climbing a few ranks on it would be read as manipulation.

No one actioned that account. But that very fear is what an intent-based system produces as a byproduct. A clean ladder can be measured. A community that no longer fears being wrong is far harder to measure.

So the question I leave behind is not whether Riot is cracking down. The question is: when a publisher holds the power to investigate, adjudicate, and enforce, how will it design its own self-correction mechanism — and will it publish it before a false-positive case large enough forces its hand?