How Riot Handles Boosting: Reading the Anti-Boost System Through Data

3:47 a.m., Brisbane time. The spreadsheet is still open.

In the fourth tab sits an account I coded A-417, tracked since February. Three hundred and twelve matches, a 47 percent win rate, a 0.94 kill-to-death ratio. Then, across eleven days, everything flipped: 78 percent wins over forty matches, kill-to-death up to 1.61, kills per round nearly doubled. One measure moved the other way. Decision time inside gunfights fell by almost 40 percent, while long-range accuracy barely budged. A player who improves aims better than before. An account driven by somebody else decides faster, because the person behind the keyboard has lived at that tempo for years.

I cannot conclude anything from that pile of rows. Neither can any viewer. Only Riot Games holds enough account-level behavioural telemetry to answer whether A-417 was boosted. That is why the company's latest Anti-Boost disclosure deserves a close read. It says nothing about professional competition, but it shows exactly where the publisher is drawing the line on cheating.

Riot Games reports having actioned 296,416 accounts for rank manipulation across VALORANT and League of Legends. The figure is pooled across both titles, with no regional split, no per-title split, and no baseline against a previous period.

A few definitions matter before going further. Boosting means a high-skill player logging into someone else's account to play ranked on their behalf and farm rating for the owner. A smurf is a secondary account created by a stronger player. Intentional deranking is deliberately losing to drop to a lower tier. Rank manipulation is the umbrella over all three, plus the buying, selling and transferring of accounts.

I have covered esports since 2026, moving from competitor to tournament organiser to media and now to data analysis for the Australian market. My job is to read the table and try not to bend it. Every number carries a story, and my task is not to ruin it.

What stands out in this disclosure is what it omits. It names no team, no player, no patch, no tournament format. The entire weight sits at the governance layer. This is a document about ladder rules, not about competitive form.

The penalty framework Riot describes has four tiers. On detection, rating points and rewards earned through cheating are cancelled, the account is returned to its original rank, and a temporary suspension follows. Repeat offences escalate in duration. Buying or selling accounts, or intentional deranking, can trigger a permanent ban. The booster's main account, along with players who frequently queue alongside them, may also be actioned.

Between those tiers sits a notable boundary. Riot states plainly that secondary accounts created and operated by the same person are normal activity. Anti-Boost targets the intent to manipulate rank, not the existence of alt accounts. The pivotal detail is that Riot is shifting from account-level detection toward match-level detection, and the company itself concedes it needs better identification of boosting signatures at match level. Read carefully, that is an admission of the current system's limits. If the method were accurate enough, there would be no need to announce an upgrade path.

The ordering of penalties also reveals priorities. The heaviest sanction, a permanent ban, is reserved for account trading and deranking, the offences with a clear money trail. The skill gap between booster and account owner is handled more lightly. Riot is attacking the payment channel, not the performance gap. It is a defensible design choice, and it tells you the service market is the real target.

The operating model is detect-then-roll-back. Points and rewards are cancelled after the conduct has already occurred. The lag between manipulation and remediation is undisclosed, but it exists by design. During that lag, every match A-417 entered degraded the experience of other players, and no mechanism returns what they lost.

A technical weakness in the disclosure itself: merging VALORANT and League of Legends into one number hides two different boosting economies. Rating-inflation pressure in a tactical shooter is not the same as in a multiplayer online battle arena. Demand for paid climbing differs by community and by season. Adding two unlike things and publishing a total is easier to read, not better analysis.

In economic terms, boosting is a grey flow. Buyers pay to climb, sellers sell skill. Strong enforcement acts as a push that raises the expected cost for both sides. But a higher expected cost does not only cut demand; it also raises price. If part of the boosting supply exits, the remainder can charge more per order. I have watched that dynamic play out in other grey markets, and it does not disappear because somebody publishes a large number.

The downstream effect matters too. A clean ladder is an input into talent-scouting pipelines. Academy systems still recruit from high-rank solo queue results, even though that has never been the only criterion. When rating is inflated, the signal thins. Nobody publishes a metric for how thin, and I do not expect one soon.

For the Oceania server, which I follow most closely, a small population gives every manipulated account a larger weight. One hundred boosted accounts on a server of tens of thousands carry a different force than one hundred on a server of millions. That is my inference, not a Riot figure, and I label it as such.

The common reading of this disclosure is that Riot is tightening the screws. The data does not say that. 296,416 is a cumulative total, not a trend. Claiming a tightening would require at least two time points and a consistent measurement definition. There is no comparative marker here, and no clear date on which the count was closed. At 39, I have learned that data hurts when it is twisted.

The numbers are self-reported by Riot and independently unaudited. Detection and adjudication both sit with the publisher, and the document describes no appeals body outside Riot. Governing authority is concentrated entirely on one side. That does not make the figure wrong, but it does mean the figure has not been cross-examined.

The joint-liability clause is the sharpest concern. "Players who frequently queue alongside them" carries no threshold. How many matches count as frequent, and over what window? If a player happens to queue with a boosted account four times in a week because their hours overlap and their rating sits nearby, are they complicit? Riot has published neither a pairing threshold nor an appeal route for that group. I once sat through nineteen match tapes just to prove whether one shot deserved to be counted as a clear chance. A process that denies punished players an equivalent right of reply will always leave a grey zone.

Intent-based standards are also hard to apply transparently. A bright-line rule banning all alt accounts would be simple to explain and simple to test. An intent-based rule depends on a judgement model, and a judgement model cannot explain itself to the person penalised. Players cannot verify in advance whether they are in violation. In traditional sport, opaque rules tend to produce contested sanctions, and I see no reason esports would escape that law.

One more paradox: if recidivism were low, an escalating penalty ladder would be unnecessary. Riot's decision to build one implies it expects a non-trivial reoffence rate. That is a reasonable assumption, and it is also an unquantified one.

The signal I will track next cycle is not the number 296,416. I am waiting for a second time point that allows comparison, for a concrete threshold attached to the joint-liability clause, and for the first public appeal to show how much weight the intent-based standard can bear. When the table speaks, the stadium must learn to stay quiet. But a table only speaks after it has survived a second cross-examination.