Riot Handled 296,416 Boosting Accounts: The Real Punch Lands on Innocent Players

Three in the morning at an internet cafe tucked deep in a back alley in the Yeonsu district of Incheon, I sat behind a boy of about nineteen. His screen showed a VALORANT ranked match. But his hands were not on his own keyboard — he was playing on an account that was not his, logging in with credentials sent by someone else over a messaging app, along with a short line of text: paid on completion, two hundred thousand per tier. Above it, the scoreboard showed the account climbing from Silver to Gold after four straight wins. The account's real owner was asleep somewhere, and when he woke up, his rank would be different.

I tell this story not to accuse a student of earning money at midnight. I tell it because around that same stretch of days, while I was tracking community forums and official channels, Riot Games published a number that made the whole industry pause: its Anti-Boost system had actioned 296,416 accounts engaged in rank manipulation, pooled across both VALORANT and League of Legends.

Three hundred thousand accounts. More than the population of several provincial cities in Asia.

But if you read that number and nod, thinking Riot is finally getting tough, you have read the wrong part. The important thing sits in the small footnote below the penalty table: players who frequently queue with a booster may also be actioned. That is the joint-liability clause. And that is the real punch.

Boosting is not the hobby of a few lazy players

Before dissecting the enforcement mechanism, it is worth stating plainly what boosting is, because most ordinary players still lump it together with the notion of a smurf or an alt account.

Boosting, in Riot's own definition, is when a high-skill player logs into someone else's account and plays ranked matches on their behalf, earning rank points for the account owner. The booster gets paid. The account owner gets the rank. Riot gets a ladder distorted at both ends: the top inflated by ranks that do not reflect real ability, the bottom drained by the very boosters climbing through it.

The difference between an alt account and boosting is technically clear but ethically blurred in the eyes of the community. A player who creates a second account, climbs on it himself, and takes his own losses — that is normal activity, explicitly protected by Riot. A Platinum player in Vietnam logging into an account for a Grandmaster in Seoul to play on his behalf — that is a violation. The same underlying act of logging into another account, but with entirely different intent, and entirely different enforcement consequences.

Why does boosting exist as an underground industry? Because rank has become a convertible form of social asset. In game, rank determines matchmaking experience, teammate quality, community reputation, and the chance of an invitation to a semi-pro roster.

Outside the game, rank determines something else: a line in a job application, a screenshot used as supporting evidence in a visa file, a tagline on a personal streaming channel. For a segment of parents and students, rank is even something to show off in group chats, like an honor-roll certificate for the digital age.

When an emblem of prestige has a price, a market appears to serve it. Boosting is the product. Account trading is the distribution channel. Intentional deranking is a technical trick to manufacture supply. That is why Riot cannot solve the problem with a simple ban wave.

Four penalty tiers and the logic of escalation

The first notable feature of Riot's documentation is a four-tier penalty ladder designed on the principle of escalation.

Tier one covers detected manipulation. The result: all ranked points and rewards obtained from fraudulent matches are cancelled, the account is returned to its original rank before the interference, and a temporary suspension is applied. This is a rollback-plus-punishment mechanism, not pure prevention.

Tier two covers repeat offenses. Ban duration increases with the number of violations. This is a detail many readers skim past, but it says something important about the publisher's assumptions: if a system needs escalation rules, the recidivism rate must be considered significant. A mechanism only matters if people fall into it a second and third time.

Tier three covers violations with a clear commercial character: buying, selling, or transferring accounts, and intentional deranking. This is where punishment crosses from temporary to permanent. Riot grouping these two behaviours in the same tier is a telling strategic signal: they read deranking to boost someone else as an economic act, not a prank.

Tier four extends enforcement beyond the directly manipulated account. The booster's main account may be actioned. And players who frequently queue with that person may also fall within scope.

This structure explains why Riot chose to publish a total figure rather than a tier-by-tier breakdown. A tiered penalty ladder needs tiered data to be evaluated. Publishing a single pooled number, with no split by severity, by title, or by region, turns information into more of a billboard than an audit report.

Intent-based standards: the right choice, at a price

This is the point I consider most important in the entire policy, and also the most widely misunderstood.

Riot does not ban alt accounts. Riot bans the intent to manipulate rank.

This is an intent-based standard, not a bright-line rule. Riot states clearly that an alt account created and operated by the same person is normal activity and will not be actioned. Anti-Boost targets the intent to interfere with the ranked system, not the existence of multiple parallel accounts.

As governance philosophy, the choice is subtle and correct. A broad ban on alt accounts would destroy entirely legitimate behaviour: players wanting to practise new agents without wrecking their main rank, players wanting to play with lower-rated friends, players wanting a separate experimental account. Distinguishing intent protects those people.

But the price of subtlety is feasibility at a scale of hundreds of thousands of accounts. An intent-based standard is far harder to explain transparently than a bright-line rule like one account, one person. When enforcement rests on inference from behavioural signals and telemetry, the possibility of error is never zero. And once a public wrongful-punishment case appears, the legitimacy of the entire standard comes under question.

In sports journalism I have long held to one principle: when the governing body is simultaneously the legislator, the investigator, and the judge, every figure it publishes must be read as self-declaration, not as a court ruling. Riot sits squarely in that model. No independent appeals body is described in the documentation. All governance authority is concentrated in the publisher's hands.

This does not mean Riot acts arbitrarily. It means players have no third channel to adjudicate when they believe they have been wronged.

The real punch: the joint-liability clause

Back to the footnote I mentioned at the start.

Riot's documentation states that beyond the directly manipulated account, the booster's main account and players who frequently queue with them may also be actioned. This is a form of extended joint liability.

As enforcement logic, the clause has a reason to exist. Boosters do not work alone in a vacuum. They often move in groups, queueing with acquaintances to secure win rates, control teammate quality, and reduce the risk of being reported. In some models, an entire cluster of accounts is operated in parallel, rotating through shared lobbies to optimise climb speed. Banning only the manipulated account would leave the operating group healthy and re-established within hours.

But the line between accomplice and bystander here is very thin.

Imagine an ordinary player who climbs the ladder every evening with a friend met through a community group. That friend is in fact operating someone else's account for a few matches, or selling boosting services under the cover of playing together. The ordinary player does not know. When the system sweeps and actions that friend, the ordinary player's account may appear on the list because the queue history matches.

Riot's documentation describes no specific queue threshold. No minimum number of matches. No defined time window. No appeal mechanism is stated. That is a large governance gap.

This is the point I want anyone reading the 296,416 figure to see: the system may be right for the majority, but its design does not exclude the possibility of sweeping in innocent players, and no second step to correct that is publicly described.

In sports history, joint-liability sanctions of this kind have always been contentious, because they shift the burden of proof from the accuser to the accused. Once you must prove you did not know, you are already at a disadvantage.

The 296,416 figure and the problem of pooling two different titles

Now for the number.

Riot published 296,416 accounts engaged in rank manipulation, pooled across VALORANT and League of Legends. No split by title. No split by region. No split by season. No baseline for comparison with a previous period.

A total is not a trend. To say tightening, you need at least two data points at two different times. To say which region is worst, you need geographic breakdown. To say which title has more boosting, you need to separate VALORANT from League of Legends.

Pooling a tactical first-person shooter with a multiplayer online battle arena into a single report is not merely a formatting matter. These two ecosystems run on different economic logic. Rank-inflation pressure, seasonal boosting demand, the semi-pro circuits that draw in talent, the community's tolerance for account buying — all differ. Pooling them loses precisely the accuracy that would make the report useful.

I have tracked ranked matches in both titles across many seasons, and the difference is plain to see. In League of Legends, boosting tends to be long-term, tied to seasonal contracts, tied to the need to maintain a rank that qualifies a player for amateur circuits and publisher talent-search programmes. In VALORANT, boosting tends to be short-term and explosive, tied to seasonal reward milestones and to the culture of rank display on social platforms.

A single figure pooling both models can be read, but cannot be analysed.

And there is one more detail: this is publisher-reported data, with no independent audit. Riot builds the detection system, operates it, and publishes its results. In finance, that is called a structural conflict of interest. In gaming, it is called Tuesday.

Match-level detection and an arms race with no finish line

Another notable part of the documentation is the expansion roadmap. Riot says it will continue scaling Anti-Boost and move toward match-level detection, based on behavioural signs of boosting within the match itself.

That is a qualitative shift. If the current system operates mainly by detecting after the fact and then rolling back points and rewards, match-level detection aims to intervene earlier, before the damage spreads across the whole ladder.

But it must be said plainly: Riot having to announce it is improving detection is an indirect admission that current methods are imperfect. Nobody upgrades a system that is already good enough.

This is the nature of an arms race with no finish line. The defending side builds behavioural signatures. The attacking side learns to break them. The defender updates. The attacker changes method. The loop does not end in victory; it ends when one side quits on cost.

There is an important cost asymmetry here. For Riot, scaling detection is a large, fixed investment that must be sustained continuously. For an individual booster, the cost of adaptation is close to zero: change accounts, change communication channels, change working hours. The resource-poor side is more tactically agile. That is the familiar paradox of every war on online fraud.

Riot says penalties for account buying and selling will be severe, potentially permanent. Economically, that is a strike at the supply side of the account market. If detection risk rises, boosting service prices must rise to compensate, and as prices rise, some marginal demand disappears. The mechanism is sound in theory.

But it is only sound if there exists a price at which demand is genuinely elastic. For people buying rank for out-of-game purposes — a job opportunity, a place in an academy, a line in a study-abroad file — demand is nearly inelastic. They will pay anything. For this group, banning accounts just means buying new ones, which feeds the very market Riot is trying to shut down.

Why boosting is a product of design, not of morality

Here I want to turn to an angle analysts usually avoid.

Boosting is widespread in VALORANT and League of Legends because these two titles turn rank into an asset that can be displayed, measured, and publicly verified. The ladder is public. Match history is public. Rank emblems are public. Everything can be screenshotted, shared, and turned into social proof.

When a measurement system acquires social value, a market appears to trade that value. This is a rule, not an individual moral failing.

Looking at other titles in the industry, communities less afflicted by boosting are usually not less afflicted because their players are more ethical. They are less afflicted because their ranks are less displayable, or because their ranking systems are harder to extrapolate from, or because their designs make individual skill harder to carry with.

Put differently: Riot can raise penalties to infinity, but demand will persist as long as rank remains a medal you can wear outside.

I made this argument in an earlier analysis on how publishers handle manipulation behaviour in competitive titles. I wrote then that once you ban a behaviour, the behaviour migrates to harder-to-detect channels rather than disappearing. What is happening with Anti-Boost confirms it. As account detection gets easier, operating groups shift to off-platform coordination, private communication channels, constant account rotation, and coordinated deranking loops to optimise speed. That is the signature of a market adapting, not a market being shut down.

The invisible hand of the digital-era agent

In football, I once wrote that player agents are the largest hidden cost of the transfer market. They create noise, they inflate prices, they distort signals about a player's true quality.

In esports, that role is played by a different class: boosting brokers, intermediary account holders, and service groups running climb operations.

They do not appear on transfer headlines. They are not in press conferences. But they create a layer of noise with exactly the same effect: they make the signal of true skill hard to read.

A scout looks at a high-rank ladder on a server and sees hundreds of accounts at the highest tiers. What share of them are real people, self-climbed, at their true level? Nobody can answer. And when the signal is noisy, the cost of talent discovery rises, and genuine talent in overlooked regions becomes easier to miss.

This is an indirect effect the Riot report never mentions, yet it matters for the whole ecosystem: a clean ladder is a prerequisite input for the amateur scouting pipeline, and a dirty ladder is an invisible tax levied on the very players with the fewest opportunities.

I have seen this in the Korean market, where youth academies screen candidates from top-tier ranked play itself. When the input data is noisy, they must rely more on internal referrals and in-person tryouts — channels that require travel costs, living costs, and relationships a child from a provincial town does not have.

Boosting does not just ruin your match tonight. It narrows the path of the children who would become professional players five years from now.

Reading history back: what I said before

I make a habit of recording my predictions and checking them against outcomes. This is the section I always place at the end of an analysis, and today is no different.

During the pandemic, when leagues stopped and I began reviewing replay matches in empty stadiums, I published a series on a phenomenon I called ghost football. My argument then: when crowds vanish, home advantage does not vanish entirely, it just changes mechanism — and home win rates in some leagues would fall noticeably. The figure I cited for one specific league was a drop from roughly forty-seven per cent to roughly thirty-two per cent. Many said I was exaggerating. International outlets later cited the series.

In 2026, before a major European final, I told the newsroom that the favoured side would lose because their coach was too cautious, and that their opponent would win even from behind. Colleagues laughed. I won a barbecue. But the bigger win was confirmation that bold, evidence-based predictions build outsized credibility compared with safe analysis.

More recently, as competitive-game publishers began publishing enforcement figures, I wrote that totals would become a communications tool more than a measurement tool, and that without a baseline every trend claim is inference. Watching how the 296,416 figure has spread with no accompanying breakdown, I see that prediction holding.

I write these lines not to make you agree with me. I write so you know there is someone out there thinking differently, and that this is fine.

Where I might be wrong

This is where I have to interrogate myself, because an analysis without self-rebuttal is just propaganda in nice formatting.

Possibility one: I am underestimating the system's real effectiveness. Riot has internal data I do not. If the actual recidivism rate is very low, then my inference from the existence of escalation rules that recidivism is high could be wrong. Escalation rules might exist purely for deterrence, not because of real demand.

Possibility two: I am exaggerating the joint-liability risk. There might be a very high internal technical threshold Riot does not publish, keeping the number of innocents affected negligible. A mechanism not publicly described is not a mechanism that does not exist.

Possibility three: I am misreading the motive. I argue boosting demand comes from the social value of rank. It may also be that most demand comes from curiosity, from wanting to play with friends, from trivial reasons with no economic character. If so, raising penalties would work better than I think.

Possibility four, and this is the one that worries me: the boosting problem may be far larger than the 296,416 figure suggests, and Riot is publishing a small slice of the picture for commercial reasons. If so, the question is not whether the system overreaches, but whether it reaches far enough.

I set out these four possibilities not to retract my argument. I set them out so you know that what I have written above is an evidence-based reading, not a closed conclusion.

Transfers, accounts, and the game of flies and honey

There is a line I often use about the football transfer market, and I find it true of the game-account market: transfers are a game of flies and honey — everyone flies in, but only those who know the exit survive.

In the account market, that exit is the ability to disappear. A banned account can be replaced with a new one within hours. A flagged identity can be abandoned and another built. The cost of rebirth is so low that punishment loses most of its weight.

This is why I argue the real battle is not in penalties but in making rebirth expensive. Permanently banning an account is meaningless if creating a new one is free. What matters is making the trace of past behaviour follow a player across new accounts — which is exactly what Riot is moving toward when it talks about match-level detection based on behavioural signatures.

To be clear: this is technically the right direction, but it opens a question about privacy and about the degree of publisher intervention in individual behaviour. When a system starts recognising you by how you play rather than by the account you log into, the line between fraud management and behavioural surveillance gets blurrier.

I am not saying Riot is crossing that line. I am saying the line has not been publicly defined.

Vietnam and Korea: two ends of the same curve

Born in Vietnam and working in Korea, I see this problem from both sides, and what I see is two markets at different phases of the same curve.

In Vietnam, boosting largely runs on individual and small-group models. Boosters are young, highly skilled players who need income and have found a market willing to pay. Transactions happen over private channels, payment goes through e-wallets, and prices fluctuate by the tier being climbed. In some communities, taking boosting work is treated as an ordinary side job, even shared openly.

In Korea, the level of organisation is markedly higher. There are professional operating groups, role division, quality control processes, and relationships with the account market and with small streaming channels. Social pressure around rank is higher, because rank in competitive titles here is tightly bound to career paths in the esports industry.

This gap creates an interesting flow: high-quality boosting demand in Korea can be served by skill supply from other regions, including Southeast Asia. A young player in Vietnam may be climbing an account in Korea, and that account's owner may be chasing a milestone to qualify for an academy tryout.

This is not a story about crime. It is a story about structural opportunity. When a talent pipeline only recruits from a certain rank band, and that rank band can be bought, the pipeline is selling lottery tickets to itself.

I wrote about this in football years ago, discussing scouting networks in developing countries. There, a European academy finds a fifteen-year-old prodigy in a poor province, and behind it is a family torn apart because nobody prepared them to deal with the system. Here, the same mechanism runs in digital form. A broker promises an academy slot. The family pays. The child is placed in an environment whose rules he does not understand. And when it all falls apart, nobody is accountable.

That does not mean academies are bad. It means a scouting pipeline without safeguards will be exploited by those who live off its loopholes.

Mic drop, I understand

At twenty-four, I once said on live television that a famous head coach bore responsibility for his team's defeat, and that his tactical setup had been obsolete for a decade.

I was heavily criticised. But afterwards I understood something: the mic drop is not the moment you attack. The mic drop is the moment you listen all the way through before speaking. The difference between a provocative writer and an irresponsible one is whether you bring evidence.

That is also how I read the 296,416 figure. I do not oppose Riot punishing boosting. I question the architecture of the punishment system, the gaps left undescribed, and the people who could be swept in without knowing why.

On an assignment years ago at an international tournament, I sat beside a veteran journalist and learned something I have carried since: the truth does not need a side, only someone willing to say it. You do not need to stand with the publisher to admit boosting is a real problem. You do not need to stand with the players to admit the enforcement system has real holes.

Both can be true at once.

What I predict, and how to verify it

This is where I stake my credibility.

Prediction one: Within one to two more disclosure cycles, Riot will publish a new figure. If it comes with a split by title or by region, the publisher is moving from communications to measurement. If it remains a single total, the goal remains communications. Verification: read exactly what is published, not what is inferred.

Prediction two: At least one publicly reported wrongful punishment case will emerge, tied to the joint-liability clause or to the intent-based standard. Verification: monitor community forums and esports media over the coming months for players claiming they were sanctioned for queueing rather than boosting.

Prediction three: Boosting operators will migrate to harder-to-detect channels, including coordinated deranking loops and off-platform communication. Verification: watch how these groups advertise services, and whether they start requiring clean-account proof or offering anti-ban guarantees.

Prediction four, and the one I most want to be right about: Another publisher will release comparable enforcement figures, and only then will anyone have a basis for comparison. A number only means something when there is another number to place beside it.

Something to take away

An empty stadium is an open book: read carefully and you will find contracts weeping and tactics cracking. An empty ladder is the same. Every inflated rank slot is a marginal note about a system buying itself off.

Riot got one thing right: it defined violation categories clearly, published a penalty ladder, and admitted it is still scaling up. That is a level of transparency many other publishers do not have.

But transparency about rules is not the same as transparency about outcomes. A pooled figure, unaudited, unsegmented, without a baseline, with no described appeals mechanism, is half of transparency. The other half — the harder half — is letting players see themselves in that system and believe that if they are swept in, there is a path to make their voice heard.

At three in the morning in the Yeonsu internet cafe, the nineteen-year-old finished his fifth match and shut down the machine. He received a small payment through an e-wallet. The account owner somewhere will wake up to a new rank and a ranked match tonight matched above his true level — meaning four other people will play a distorted game they never chose.

Riot can ban the boy. Riot can roll back the account owner's rank. Riot can scale the system and catch hundreds of thousands more cases.

But as long as being good at the game remains a ticket you can buy rather than a road you must walk, there will always be someone sitting in the dark, logging into someone else's account, finishing the match that someone else paid not to play.