Four Penalty Tiers and the Joint-Liability Trap: Dissecting Riot's Anti-Boost System

There is a moment every ranked climber has tasted: an opponent who shoots like a Diamond player on an account sitting in Silver, and you have no way to prove anything. You just lost a match that was rigged before it began. That feeling — not defeat, but helplessness — erodes a ladder faster than any loss. Riot Games calls what sits behind that feeling rank manipulation. From late last year to now, they published a number that made the community stop: 296,416 accounts flagged for rank manipulation across VALORANT and League of Legends, handled by a system named Anti-Boost. Based on my experience tracking ranked matches and major tournaments, the interesting part of this number is not its size. It is that this is a claim from the referee itself. And like any self-reported claim, it deserves to be dissected with a scalpel, not applauded. To understand why Riot built a dedicated system for this, you have to look at the economics of boosting. Boosting is a high-skill player logging into someone else's account to climb the ladder on their behalf. This is not a lazy individual's impulse. It is a market with supply, demand, prices and middlemen. Buyers pay for a rank they cannot reach themselves. Sellers offer their skill as a service. In between sits a network of account brokers operating parallel to the official ranked system. This market drains the very thing Riot sells to players: the feeling that rank reflects true ability. And once rank stops reflecting ability, the entire talent-scouting ecosystem behind it starts to shake. Notably, Riot defines the violation narrowly. It targets the intent to manipulate rank, not the existence of alt accounts. Riot states clearly: a self-created, self-operated alt account is normal activity. What Anti-Boost hunts is intent to manipulate, not multi-account ownership. The system reads behavior, not ownership. As someone used to dissecting tactical maps, I recognize this as the kind of rule design a coach uses when picking a defensive block: instead of banning an entire zone, you lock only the lane your opponent will run through. It is more effective and does not break legitimate play, but it demands far greater precision in reading intent. That is why this system is both elegant and fragile. Anti-Boost runs on a four-tier penalty ladder, and dissecting it reveals Riot's design philosophy more clearly than any press release. Tier one: upon detection, ranked points and rewards earned through cheating are cancelled, the account is returned to its original rank, plus a temporary suspension. This is rollback logic — the system strips what was created by cheating, then punishes. Tier two: repeat offenses escalate ban duration. This matters psychologically. The system does not just punish the act, it punishes repetition. Tier three: buying, selling or transferring accounts, or intentional deranking, can lead to a permanent ban. Tier four, the most controversial: associated parties may also be actioned. The booster's main account, and even teammates who frequently queue with them, fall within scope. Tier four is where I want to linger longest. It introduces a principle of joint liability — something traditional sports apply only when there is clear evidence of complicity. Here, the criterion is "frequently plays with". That is a broad standard. It is like a football league punishing players merely for frequently pairing with someone suspected of match-fixing. Reasonable as deterrence, but it opens a very large gray zone. The second point is detection cost. In economics, when the cost of tracing a violation exceeds the profit from enforcing it, the system tacitly tolerates that behavior. Boosting sits exactly at this intersection: buyers and sellers trade off-platform, use private channels, and the only thing that surfaces is anomaly in match data. Riot has no courtroom for interrogation. It has signals. So Anti-Boost is essentially a system reading behavioral footprints — tempo, win-loss patterns, the skill correlation between account and in-match performance. It does not prove guilt, it infers guilt from probability. This is where the nature of this law is exposed: a law based on intent, enforced by probability, with the lawmaker acting as both prosecutor and judge, and no independent appeal mechanism described. The map is only correct until the ball lands. And in this case, there is no ball to compare against — only an algorithm and a claim. Riot also states it will expand Anti-Boost, adding the ability to detect boosting signs at match level rather than only account level. This is a resolution upgrade. Instead of looking at an account's long history, the system reads signals within each match. Technically, this makes sense: if boosters grow more sophisticated at disguising account history, in-match reading is an inevitable response. But it also means the intervention threshold will drop, and the number of gray-zone cases will rise. This is where I must say plainly something the community rarely wants to hear: the more sensitive a defense system, the higher the false-positive rate. No exceptions. The question is not whether someone will be wrongly punished, but how many, and whether they have a path to appeal. Behind this layer of rules lies an under-discussed layer of value. Riot publishes the total number of actioned accounts as a reputational signal — a way of telling players and investors that the ladder is being actively managed. For a publisher owning the entire ecosystem from game to pro circuit, the ability to self-enforce integrity is a competitive advantage against titles perceived as laxer. But precisely because there is no third-party verification, the 296,416 figure is a claim, not an independently verified proof. Another notable point is that Riot pools VALORANT and League of Legends into a single number. Statistically, this is noisy presentation. The two titles have different boosting economies: rank pressure, market demand and account characteristics differ. Pooling hides each side's own dynamics and makes any trend conclusion more fragile. The counterintuitive angle here is not that Riot is cracking down on boosting — everyone agrees that should happen. It is that the way they do it is creating a new risk layer no one puts on the scale. First, the joint-liability clause for teammates who frequently queue together is an invisible trap. A normal player could unknowingly climb alongside a booster for many matches, never knowing the other is boosting, and suddenly land on the actioned list. No pairing threshold is stated, no appeal mechanism. That is a governance gap, not a minor detail. Second, the 296,416 figure is self-reported data, unaudited. It gives a total, not a trend. Without a prior-period baseline, you cannot say Riot is "tightening" — only that they actioned that many accounts over an unspecified window. This is the kind of reasoning I learned from running simulations: a single data point does not make a line. Third, the biggest trap is the asymmetry between detection and evasion. The system learns to recognize, boosters learn to disguise. This race has no finish line. And when a law is enforced by probability without a courtroom, every time the system grows more sensitive, more innocent players get swept into the gray zone. The greatest victories are often woven from a trap no one sees. Here, that trap is not for boosters, but for honest players standing in the wrong place. Pitch and Map are not opposites; they are two ways of drawing the same trap. Here, Riot is drawing a trap that is reasonable in aim but blurry in boundary. Every arena has a map; the winner is the one who reads the map before the ball rolls. What the community should watch is not Riot's next published number, but whether they build a real appeal mechanism and a transparent threshold for the joint-liability clause. A clean ladder is not built on fear, but on trust that the rules are applied fairly. And that trust, once eroded, is something no algorithm can restore.