Anti-Boost's Four-Tier Penalty Ladder: How Riot Is Re-Reading the Ranked Map of VALORANT and League of Legends

In a VALORANT ranked match I rewatched one night late last year, one player made me stop at the seventh minute. Not because of a miraculous gunfight, but because of an implausible consistency: aim angle, weapon-swap rhythm, peek timing — all identical across rounds, as if the same hand were repeating a memorized drill. I rewound three times. By the third, I understood: the person behind that account was most likely not its owner.

The real story is not the booster. It is the system standing behind them. Riot Games, in its latest Anti-Boost disclosure — an automated ranked-manipulation enforcement system — reported handling 296,416 accounts with rank-manipulation behavior across VALORANT and League of Legends. Four penalty tiers. A clearly drawn safe harbor. And one clause at the final tier that I regard as the largest blind spot in the whole system. The greatest victories are woven from a trap nobody sees — this trap weaves not a victory, but a risk.

A gray market built on stolen trust

Ranked standing in VALORANT and League of Legends is not a number for bragging. It is a social currency: it decides who gets matched with whom, who gets invited to semi-pro rosters, who draws the eye of organizations during scouting cycles. When a currency has value, a black market emerges. When a black market emerges, middlemen appear to take the spread.

Boosting is the entry point of that chain. A high-skill player logs into someone else's account, plays ranked on their behalf, and brings points back to the owner. Around it sits a secondary ecosystem: account buying, selling and transfer; intentional deranking to face weaker opponents; and smurf accounts used as climbing tools. Riot groups these behaviors under "rank manipulation," and Anti-Boost is the machine responsible for detecting and penalizing them.

I once simulated 100 matches during the COVID season using Football Manager to find how a crisis bends tactical behavior. That experience taught me one thing: to understand a system, look at its rewards and punishments before its statements. With Anti-Boost, the reward is a clean ladder. The punishment is four tiers. And those four tiers are what deserve dissection.

Four penalty tiers — read as a bracket

The first tier handles detected manipulation. Consequence: ranked points and rewards earned from cheating are cancelled, the account is returned to its original rank, with a time-limited suspension. This is the lightest preventive tier — essentially a rollback plus a warning.

The second tier handles repeat offenses. Ban duration escalates with the number of violations. The very existence of an escalation mechanism is information: if the recidivism rate were negligible, the ladder would be unnecessary. Riot designed an escalation ladder, which implies they anticipate a population of players who return to old behavior after punishment.

The third tier is technically the heaviest: buying, selling or transferring accounts, or intentional deranking, may lead to a permanent ban. This is a notable policy design point. Riot does not equate all manipulation behavior. The heaviest penalty is reserved for commercially motivated, clearly intentional conduct — account sales, account purchases, self-deranking for profit. In other words, they prioritize hitting the supply side of the black market rather than only the symptoms.

The fourth tier extends liability to related parties. The booster's main account may be actioned, and players who frequently queue with that booster may also fall under enforcement. This is where I want to pause longest, and I will return to it later.

Alongside these four tiers, Riot draws a very clear safe harbor: alt accounts created and operated by the player themselves are normal activity. Anti-Boost targets the intent to manipulate rank, not the existence of an alt account. This is an intent-based standard, not a hard technical line.

As an observer, I think intent-based standards are philosophically correct but difficult to enforce. A hard line — banning all alt accounts, for instance — would be transparent but would strangle entirely legitimate behavior. An intent-based standard protects ordinary players but places the entire burden of proof on the detection system. When a system must infer intent from behavioral signals, error is unavoidable.

This is where a principle I drew from my own writing applies: the more a system relies on inferring intent, the greater the cost of a single misjudgment. Permanently banning a legitimate account is not a minor technical error. It is a trust error, and trust in a ranked community is an asset that cannot be returned.

The economics of boosting

To understand why this market exists, view it through the eyes of an economy player in CS. In Counter-Strike, every round is a decision between risk and return: buy or save, force or concede. The booster runs a similar calculation, except the currency is real money and the price is the probability of detection.

Anti-Boost operates as a mechanism that raises the expected cost of violation. When detection probability rises, service prices must rise to compensate for risk, or margins erode. That is why the commercially oriented penalties at the third tier — permanent bans for account buyers and sellers — matter more economically than short suspensions. They strike directly at the asset the black market uses as its unit of exchange.

But there is a paradox the report itself does not resolve. If the black market survives after 296,416 accounts were actioned, then either detection cost remains too low, or demand is too high to be eliminated by administrative measures. Demand for rank does not vanish when an account is banned. It shifts to another account, another seller, a more closed channel.

I once wrote about this on the Sân Cỏ & Bản Đồ blog, comparing how traditional sports leagues handle match-fixing. Football federations cannot erase gambling. They only make fixing more expensive, riskier, and socially branded as shameful. Riot runs the same formula, differing only in that it operates at the digital layer, where behavioral evidence can be collected automatically rather than investigated by hand.

The trap at the final tier: joint liability

Now back to the clause that troubles me most: players who frequently queue with a booster may also be actioned.

Logically, this makes sense. Boosters rarely act in a vacuum; they often play in groups to raise win probability and cut climbing time. If only the manipulated account is banned, the booster keeps their main account and keeps earning. Extending liability to related parties closes that gap.

But this is precisely the blind spot. An ordinary player may accidentally queue with a booster across several consecutive matches — entirely normal at high ranks, where the community is small and re-pairing with the same partner is frequent. The report states no tolerance threshold, no minimum match count, and no appeal mechanism. Without those, the line between "accomplice" and "incidental teammate" blurs, and the burden of proving innocence falls on the player.

I call this structural false-positive risk. It is structural because it does not come from an algorithm error but from the rule design itself — extending liability without a matching protective mechanism. In football, when a club is punished for fielding an ineligible player, a clear registration record always exists for cross-checking. Here, the equivalent of a registration record is a queue-pairing pattern — and patterns are far easier to misread than a signature on paper.

It bears adding that every penalty Riot publishes is detected, judged and enforced by Riot itself. No independent appeals body is named in the report. That concentration of authority is characteristic of publisher-run governance — efficient in speed, but lacking an external cross-check layer. In traditional sports, independent sports arbitration courts exist for exactly this purpose. In online ranked play, that role is currently absent.

The nature of the data also needs stating. 296,416 is a cumulative total self-reported by Riot, without independent audit. I have no technical reason to doubt it, but in analysis I always distinguish two data types: third-party verified data, and self-reported data. The latter is usable, but must be used correctly.

The contrarian angle

The report asserts Riot is tightening enforcement. Read closely, the cited data is a single total. A total does not indicate a trend. To prove "tightening," you need at least two time-based comparison points — for example, accounts actioned this quarter versus last quarter. Riot offers a photograph and describes it as a film. That is an inference, not a fact.

This matters because it changes how the whole message reads. Without a trend, the claim that "Anti-Boost is getting stronger" loses its data footing, and the rest is merely a statement of intent. Intent still has value — it shows Riot is ready to keep investing in scaling Anti-Boost and adding match-level boosting-sign detection. But it does not prove current effectiveness.

The second contrarian angle concerns asymmetry. A detection engine must be right in every case to protect trust, while a violator only needs to be right once to escape. In this contest, the attacker always holds the adaptation-speed advantage. The report's admission that detection of boosting signs is still being improved is a gentle confession that current methods are imperfect.

Every arena has a map; the winner reads the map before the ball rolls. In the anti-boosting story, the map is not the detection algorithm — it is the incentive structure that makes players want to buy rank. Riot is reading that map from the supply side. The demand side remains untouched.

Closing

Pitch and map are not opposites; they are two ways of drawing the same trap. Football learned long ago that the credibility of a table is protected not by moral statements, but by registration records, independent courts, and penalties that make violation economically pointless. Esports walks the same road, only faster — and in moving faster, it risks skipping the cross-check layer that traditional sports took decades to build.

I do not doubt that Anti-Boost needs to exist. I question the joint-liability clause at the final tier, and whether a system that self-detects, self-judges and self-enforces can convince those it inadvertently touches. The map is only true until the ball lands. And in the match between publisher and black market, the ball is still rolling.