Riot Sanctions 296,416 Boosting Accounts: The Anti-Boost System and Three Unclosed Gaps

On November 12, I reopened fourteen consecutive ranked replays of a VALORANT account on the Vietnam server inside an internet cafe in Thu Dau Mot, Binh Duong. The owner sat at Diamond 3. The most recent three days: thirteen wins in fourteen matches, an average K/D of 1.87, a 42% headshot rate, and a 61% duel-win rate. Three weeks earlier, that same account: forty matches, eighteen wins, a 0.91 K/D, and 19% headshots. No patch changed agent balance enough to explain that gap. No hardware upgrade explained that gap. Someone else sat in that chair. By Riot Games' own behavioural definition, that is boosting. And by the publisher's official disclosure, its Anti-Boost system has actioned 296,416 accounts across both VALORANT and League of Legends. To understand why that number matters to Vietnamese players, it needs context. Boosting is not new. In football, the equivalent is a ringer — a strong player wearing the shirt of an amateur side to guarantee promotion. In esports, boosting is a high-skill player logging into someone else's account and playing on their behalf, banking rank points for the account owner. The payer receives a rank that does not match real ability. The earner gets income outside any formal competitive structure. And the ladder — the instrument meant to measure real skill — turns into a marketplace. Riot Games has not stayed silent on that market. The Anti-Boost system classifies four categories of violation: playing ranked on another person's account; buying, selling, or transferring accounts; intentional deranking; and climbing through a low-skill alt. The most important detail sits in how Riot draws the line: a self-created, self-operated alt is still considered normal activity. Anti-Boost targets the intent to manipulate rank, not the existence of an alt account. That is a narrow standard. And its narrowness raises three questions this piece walks through one at a time: does the penalty ladder actually deter? Does the joint-liability clause covering frequently paired teammates exceed a reasonable scope? And what figure is Riot using to grade its own effectiveness? Four penalty tiers and the latency trap Anti-Boost runs on a four-tier escalating ladder. Tier one: on detection of manipulation, ranked points and rewards derived from cheating are cancelled; the account is returned to its pre-manipulation rank; and the account is temporarily suspended. Tier two: repeat offences lengthen the ban. Tier three: buying or selling accounts, or intentional deranking, can trigger a permanent ban. Tier four: related parties — the booster's main account and frequently paired teammates — can also be actioned. It reads reasonably. But there is one detail skimmers miss: point cancellation and rank rollback are reactive measures applied after detection, not preventive controls applied before. That implies a latency window between the manipulation and the remedy. Inside that window, the buyer has already received what they paid for — a virtual rank — and may already have used it to enter community tournaments, amateur tryouts, or unofficial betting markets that treat rank as a reference point. Don't rush to look at the score; look at how they move without the ball. Here, moving without the ball is the number of matches contaminated before the system intervenes. In traditional anti-doping, latency is offset by retroactive sanction: medals stripped, records erased, results amended. With boosting, Riot does something similar — cancelling points, rolling back rank — but cannot restore the matches the boosted account already distorted for other players. Four real teammates, every match, won or lost on a fraudulent variable. Nine opponents lost points to someone who did not belong at that rank. An empire does not collapse overnight; it collapses from the moment it believes it is an empire. A ranked ladder does not collapse because of one boosted account, but because hundreds of thousands stack up — and because of the belief that rank can be bought. That belief spreads faster than any patch. Joint liability: shield or dragnet? Of the four tiers, the fourth draws the most dispute. Riot states it may action teammates who regularly queue with a boosted account. The reasonable argument behind it: some boosters work in groups, queuing with a few familiar accounts to raise their win rate; if only the boosted account is banned, the organiser stays outside the internal legal perimeter. So Riot widens the net. That is exactly where the crack runs. The frequently paired with criterion has no published quantitative threshold. How many shared matches constitute complicity? Five? Twenty? And among those who queue alongside, who actually knew they were playing with a boosted account, and who simply logged in at the right time, in the right queue, with the right friends? Riot publishes no threshold and describes no appeal mechanism for those hit indirectly. That is a governance gap, not a technical bug. For Vietnamese players — where internet-cafe culture and shared accounts remain common — this risk is not theoretical. I have watched five friends sit at a row of machines in a Binh Duong cafe, logging in on each other's accounts because someone forgot a password or had not finished setup. Most of those cases are harmless. But under a standard built on queue behaviour and telemetry signals, a harmless group can land on the same list as an organised boosting ring. At that point the question is no longer whether the person is guilty, but whether they have the evidence to defend themselves. The asymmetry between detection and evasion Riot acknowledges it is still improving its ability to detect signs of boosting at the match level, not only the account level. That is the right direction. It is also an implicit admission that the current detection method is imperfect. Data does not create a revolution; it exposes who is running on instinct. In the anti-boost fight, data exposes an uncomfortable truth: detection always trails evasion by one step. When Riot closes one method, the boosting market finds another. When Riot monitors in-match signals, boosters shift to off-platform communication. When Riot tightens account trading, the market shifts to hourly or per-match account rentals that are harder to trace. This is not unique to Riot. It is the structure of every anti-fraud fight. The question is not whether evasion exists, but which way the ratio between detection cost and evasion cost tilts. If evasion cost is systematically lower than detection cost, then 296,416 actioned accounts may be the visible tip of a growing iceberg rather than proof the iceberg is melting. An empty stadium, yet the number shouts louder than any crowd. Here, the stadium is the ranked queue, and the shouting number is the share of actioned accounts against total active accounts. Riot publishes the numerator. It does not publish the denominator. The Vietnamese context: internet cafes, shared accounts, and the VCS When Riot publishes global figures, Vietnamese players often read them as an event happening somewhere far away. But Vietnam's boosting market has its own features worth facing directly. First, infrastructure. Internet connectivity in Vietnam's major cities is stable enough to play VALORANT and League of Legends at competitive ping. That means a booster in Vietnam can take orders from clients anywhere, provided they have an account and a connection. Geography is no longer a barrier. Second, shared-account culture. In many internet cafes, friends sharing an account is ordinary. The line between harmless sharing and organised boosting is sometimes faint. Riot draws the line by intent, but intent is hard to prove and hard to verify from the outside. Third, the scouting pipeline. The VCS and Vietnam's regional VALORANT circuits recruit from high rank in solo queue. If the ladder is contaminated by boosted accounts, the scouting signal is distorted. A scout looking at Challenger cannot be certain who sits behind the account, or how good they really are. That is an indirect but serious consequence: it does not destroy a single match, it destroys an entire talent-discovery channel. Their failure does not come from bad luck; it comes from faulty design. If the scouting system rests on a ladder that can be bought, the faulty design is over-reliance on a manipulable index. A comparison with anti-doping in traditional sport Set Anti-Boost beside the anti-doping systems of traditional sports federations and three similarities and one major difference emerge. First similarity: both rest on behavioural and intent-based standards, not merely outcomes. You are not sanctioned for having a banned substance in your body; you are sanctioned for putting it there. Likewise, you are not sanctioned for having an alt; you are sanctioned for using one to manipulate rank. Second similarity: both use escalating penalties. Repeat offences are punished harder than the first. That is a universal principle of disciplinary governance. Third similarity: both include joint-liability clauses. In doping, that means coaches, team doctors, or managers who knew and did not report. In boosting, it means frequently paired teammates. The major difference lies in the appeal mechanism. The international anti-doping system has an independent Court of Arbitration for Sport, where athletes can appeal and be heard by a third party. Riot has no equivalent independent adjudicatory body. The publisher investigates, prosecutes, and rules. That concentration of authority is efficient in speed terms, but it places the entire burden of proof on the sanctioned party. This is where I start looking for the crack when everything seems too stable. The faster a system works, the more it needs procedural transparency. Otherwise every correct sanction is recorded, while every wrong sanction is buried — until one high-profile case erupts and drags a chain reaction behind it. The scouting pipeline and the downstream amateur cost At thirty-six, I no longer only write hot takes. I care about what remains after the piece ends. And what remains in the anti-boost fight is the quality of the scouting pipeline. A clean ladder has direct economic value to esports organisations. Academies, youth teams, and scouts use high-rank solo queue as an entry filter. When the filter is contaminated, screening costs rise. Organisations must hire more reviewers, run more trial scrims, and invest more time per prospect. That is a real, measurable cost, and it lands on small organisations — the ones with the fewest resources. There is a rarely stated paradox: boosters are often players skilful enough to be considered for scouting. They chose a fast-earning path over a professional career path. For some, boosting is primary income, more attractive than an academy wage or an amateur player's living standard. If the system offers no alternative path, it is competing directly with its own scouting pipeline. Glory is only the top of the tree; the root is who dares take responsibility. The root here is the question: who is responsible for creating an environment where good players choose boosting over competition? Riot can ban 296,416 accounts. But banning accounts does not build an alternative career path. The reverse view: where I might be wrong I have laid out a fair amount of criticism of Anti-Boost. At this point I have to interrogate myself. First, I am reading figures from a single source: Riot's own disclosure. Across eighteen years of covering sport, I remind myself that self-reported figures are not verified figures. If a third party — an independent research body, a game-data analytics platform — publishes cross-checked numbers, my conclusions could change. Second, I may be underrating Riot's rate of improvement. A match-level detection system, correctly deployed, could be far more precise than account-level detection. If Riot solves that problem, much of my latency critique loses its footing. Third, I may be conflating two kinds of risk: the risk of a system being abused, and the risk of a system being opposed. A system opposed for being too strict can still be the right system. A system abused for being too loose is the wrong one. I lean toward Riot sitting in between, and I should take that possibility more seriously instead of only hunting cracks. Fourth, I may have ignored the fact that any large-scale governance system must choose between speed and precision. No system achieves both simultaneously in its early phase. Riot chose speed first, and that may be sensible for the current phase — provided it shifts toward precision before trust erodes. Closing with a verifiable prediction I am not writing to summarise. I am writing to set a checkpoint. My prediction: within the next six to nine months, Riot will publish an Anti-Boost update that includes a baseline comparison — one number set beside another, not a number standing alone. If that happens, Riot is moving from a speed phase to a transparency phase. If it does not, and if we still get only cumulative totals with no comparative anchor, I will continue to treat the tightening claim as marketing rather than a measured outcome. And I will track one specific case: a legitimate account actioned for queueing with a booster, publicly documented, with a response from Riot. How Riot handles that case will say more than 296,416 numbers added together. Because in the end, the quality of a governance system is not measured by how many people it bans, but by whether it can correct itself when it bans wrongly.