Riot Games and the Anti-Boost System: When 296,416 Accounts Were Flagged and the Blurred Line of Fairness

My rented room in Busan sits at the bottom of a small hill, where each night I can hear the last subway train pass before the city falls silent. That night I stayed awake until three in the morning, not to watch any final, but to reread three weeks of notes tracking an anonymous ranked account on the VALORANT ladder. The player climbed from Silver to Gold in seven days — a pace anyone who has played long enough knows is abnormal. I recorded the timing of each match, its duration, and the gaps between them. Some nights the account played for ten straight hours. Other nights it vanished and returned with a completely different style.

No one was caught. No one spoke up. The matches went on, the points kept accruing, and the players actually in those games — the ones who lost — never knew they had just been beaten by another hand holding the mouse somewhere else.

That was the moment I understood why I pursue this subject. The pitch never sleeps; people simply choose to look away. In esports, that "pitch" is an invisible ladder, where hundreds of thousands of matches happen daily with no one sitting down to write the minutes. No stands, no linesmen, no camera recording the exact moment an account changes hands.

Context: A number that appeared amid silence

Earlier this season, Riot Games published a report on its Anti-Boost system — the automated tool it uses to detect and penalize boosting and rank manipulation across both VALORANT and League of Legends. The number announced made me reread it three times: 296,416 accounts showing signs of rank manipulation.

Not because the number is large — though it genuinely is. But because of how it was presented. Riot pooled two titles into a single report, with no regional split, no per-title split, and no comparison baseline against a prior period. As someone used to reading women's player stat sheets for six years, I know a pooled figure always tells less than its surface suggests. But it also says something else: Riot's leadership decided to disclose what publishers typically keep private.

What is Anti-Boost, according to Riot's own description? It is a system designed to recognize four categories of behavior. First is boosting — a high-skill player logging into another person's account to play on their behalf, helping the owner climb. Second is buying, selling, and transferring accounts. Third is intentional deranking — deliberately losing to lower one's own rank. Fourth is climbing with the assistance of a smurf account.

What made me pause longest is a small point sitting in the middle of the rules. Riot states clearly: alt accounts that players create and operate themselves are normal activity. Anti-Boost targets the intent to manipulate rank, not the existence of alt accounts. It is a narrow line, and to me, it carries more meaning than every other number in the report.

Core: Four penalty tiers and a fragile line

When I arranged Riot's disclosures into a structured table, the penalty system emerged as four distinct tiers. The Anti-Boost system operates across four tiers: cancelled points and temporary rank rollback, escalating account bans on repeat offense, permanent bans for account trading and intentional deranking, and joint liability extended to the booster's main account along with teammates who frequently queue with them.

Tier one is basic handling. When the system detects manipulation, ranked points and rewards gained through cheating are cancelled. The account is returned to its original rank before the manipulation. A temporary suspension accompanies this. It is the lightest measure, and it is more corrective than punitive — the goal is to restore the state to what it was before the violation.

Tier two is escalation. Repeat offense brings progressively longer account bans. This detail matters more than its surface suggests. If the recidivism rate were low, an escalation mechanism would be unnecessary. That Riot designed an escalation ladder shows it anticipates people returning after being penalized. It is an implicit admission of the limits of deterrence — and an admission that a single ban is not enough to stop someone from resuming old behavior.

Tier three is the heaviest measure: permanent ban. Riot reserves this penalty for two commercially motivated behaviors — account trading and intentional deranking. Both carry clear economic motives. A buyer pays for a rank they cannot climb to themselves. A seller profits from it. Intentional deranking usually serves the goal of creating low-rank accounts for resale or easier future climbing. These behaviors sit at the intersection of cheating and business, and Riot handles them at the heaviest level.

Tier four is the most controversial, and I want to devote most of the remaining analysis to it: joint liability.

As described, when an account is identified as a booster, the system may also act on the person's main account and on players who frequently queue with them. On design grounds, this has its own logic. A lone boosting account does not exist in a vacuum. If it always travels with a fixed group, punishing only one account leaves an obvious gap. The violator just creates a new account and continues.

But the other side is just as clear. I once sat in the waiting room of a youth women's tournament, listening to two athletes talk about why they had been flagged by an automated system when they had done nothing wrong. That story is not unique. Over years of watching ranked systems, I have learned that when a system uses behavioral signals — rather than direct proof of account ownership — to determine violations, the possibility of error always exists at some level. With a joint-liability clause, the person swept in might be a random player from one evening, or a friend who queues together a few matches a week.

This is where I want to linger longest. The "frequently queue with them" clause has no specific match threshold. It comes with no publicly described appeal mechanism. In such a system, an honest player can be penalized with no clear way to prove innocence. The joint-liability clause for players who frequently queue with a booster is the single largest false-positive risk in the entire system, and it comes with no appeal mechanism or queue threshold of any kind.

Anti-Boost's core mechanism detects by intent, rather than banning alt accounts outright. This is a deliberate design choice. Riot understands that many players keep multiple accounts for legitimate reasons — they want to practice a new role without losing their main rank, or simply want to play with friends at a lower level. Banning all alt accounts would cost Riot a significant share of loyal users merely for using a valid feature.

But an intent-based choice has its own price. A line drawn by intent is harder to apply consistently than a bright-line rule. If two accounts show nearly identical behavior patterns but one is penalized and the other is not, players will believe the system is unfair — regardless of whether that is actually true. In ranked systems, player trust in fairness matters no less than fairness itself.

I have watched ranked games for years, and what stands out here is that Riot is trying to answer the hardest question of any anti-cheat system: how to distinguish an ordinary player from an exploiter when both use the same tool and display the same surface behavior. There is no perfect answer. Only different trade-offs.

Riot trades in favor of protecting multi-account players while accepting false-positive risk at the liability tier. From the perspective of someone who writes about fairness in sport, I find this an honest choice. It is not empty moralizing, but an admission that the system has limits and that Riot accepts those limits rather than expanding its power without bound.

Another notable factor is the system's reactive nature. Cancelled points, clawed-back rewards — all happen after manipulation has occurred. There is always a lag between the violation and the response. During that lag, matches have been played, players have lost, and points have been taken from others — even if only temporarily. The later correction cannot return the time an honest player lost, nor the feeling of being cheated they experienced in that very match.

Riot says the system will keep expanding and will add match-level detection of boosting signals. This is a reasonable step, but it is also an admission that current methods are insufficient. If they were sufficient, expansion would not be needed. The race between detector and evader has no finish line, and Riot seems to understand this better than most.

Contrarian angle: The number cannot prove what people want it to prove

Most articles on this subject end with a line like "Riot is cracking down harder than ever." I do not write that way, and here is why.

The figure 296,416 is a cumulative number. It has no comparison baseline. No prior-period figure to check against, no quarterly split, no per-title or per-region split. Without a baseline, a cumulative number cannot prove an upward trend. It only tells the total, not the rate.

I have been in this profession long enough to know that a number without a baseline is easily misused. It can be read as evidence of a forceful campaign, or as evidence of a persistent problem, depending on what the reader wants to see. Neither reading has a data basis for confirmation. A number without a baseline is not proof of a trend; it is only proof of existence.

Moreover, this number is self-reported by Riot. No independent auditor verifies it. This does not mean the number is wrong — but it means the number should be read as a publisher's claim, not as an externally verified fact. By the standard I set for every article, an unverified data point must always be marked as unverified.

There is one more point worth noting about how the report is presented. Pooling VALORANT — a tactical shooter — with League of Legends — a multiplayer online battle arena — into a single number is an analytically odd choice. The two titles have different ladder structures, different community cultures, and different boosting demand. A League of Legends player climbs differently from a VALORANT player, because the weight of individual skill differs between the two titles. The pooling strips away the ability to analyze each title separately, turning a dataset that could be highly valuable into a less informative number.

I do not believe Riot deliberately hides anything. But I do believe this presentation reflects a reality: Riot's governance system operates as a single global bloc, while boosting is a distinctly regional issue. Boosting demand depends on whether a region has a developed account market, and that varies widely around the world.

There is another angle I want to bring in — one few articles on this subject mention. Boosting does not only affect ranked players' experience. It also affects the scouting system. In many sports, and esports is no exception, teams and academies use high ladder rank as a signal to find talent. If the ladder is manipulated, that signal becomes noisy. A player with high rank through boosting can displace a genuinely talented player who has not yet been discovered. This is an indirect effect, and I mark it clearly as my inference, not a fact stated in Riot's report.

From my experience covering youth women's football in Korea, I know scouting systems depend heavily on signals. When signals become noisy, the ones who suffer are always those without a voice — those without an agent, without a network, without the ability to promote themselves. In esports, that is often young players from overlooked regions. This is why I see boosting not merely as a publisher's technical problem, but as a question of fairness in how opportunity is distributed.

What is changing

I return to that Busan night. What became of the account I tracked climbing from Silver to Gold in seven days, I do not know. Perhaps it was penalized. Perhaps it is still there, just under a different name, in a different match, with a different player on the other side of the screen losing without understanding why.

What I do know is that Riot chose not to stay silent. Publishing a number, however imperfect, is a meaningful act. It turns the issue from rumor into a figure that can be debated, checked, and tracked over time. That is the first step of any transparent governance process.

Where people wait for miracles, I learned to write with facts. The fact here is a system that exists, that operates, that has limits, and that is expanding. No miracle erases boosting overnight. There are only penalty tiers built gradually, numbers disclosed gradually, and lines adjusted gradually.

In six years writing about women's sport, I have learned that fairness does not come from one grand decision, but from many small decisions maintained steadily. Anti-Boost is such a chain of small decisions. It is not yet perfect. But it exists, and that matters more than whether it is perfect immediately.

What I want to see next is not a larger number. It is a number with a baseline. A specific queue threshold for the joint-liability clause. An appeal mechanism for honest players swept in. And a breakdown by title and by region, so that people doing work like mine can analyze rather than merely quote.

The ladder is not a pitch with stands. No one applauds when a boosting account is banned. But something similar is happening: a system learning to protect itself, and a community learning to trust that system. Esports does not need a pitch, but it still needs storytellers willing to keep the fire alive.