Riot Needs to Change the Worlds Format If It Wants Fearless Draft to Truly Work

When Riot Games introduced Fearless Draft in 2026, it stated plainly that the goal was more champion pick/ban diversity. Look at the Worlds 2026 Swiss stage and a blunt contradiction appears: the first three rounds are best-of-one, and Fearless Draft has literally zero mechanical effect in a best-of-one. The mechanism is switched off in exactly the matches that make up most of the stage, and switches on only where the tournament's stakes are highest. This is not a vague complaint — it is a precisely locatable structural flaw. And fixing it requires changing the format, not the rule. First, a correction that most commentary gets wrong: Fearless Draft is not a patch. It does not touch champion stats, items or maps. It changes the economics of the draft. A champion picked in one game of a series cannot be spent again in later games of that series. The correct analytical object is therefore a 'champion-pool tax', not a meta shift. A patch changes how strong a player is; Fearless Draft changes how much resource a team has left. Now the arithmetic at the centre of the debate. Under the standard implementation, only picked champions are locked. A best-of-one contains no prior games, so carry-over is zero — it is an ordinary draft. By contrast, before game three of a best-of-three, 20 champions are already locked; before a deciding game five of a best-of-five, 40 are gone, roughly 30 percent of a ~170-champion pool. The mechanism is binary, not gradual: fully inert in BO1, fully active in BO3 and BO5. So saying Fearless Draft is 'underused' in the Swiss stage is imprecise. It is switched off for exactly the rounds that use BO1. The shape of the Worlds Swiss stage exposes the problem. Rounds one to three are BO1; rounds four and five are BO3 deciders. On the standard 16-team assumption, 24 of 40 Swiss matches — about 60 percent — are single games. The mechanism's benefit therefore lands in the lowest-stakes segment of the tournament, while its cost lands in the highest-stakes one. A reform whose entire purpose was to show viewers champion diversity is half-neutralised by the publisher's own format. The most cited example in this debate is TES. In three Swiss games, top laner 369 used only two champions, Ambessa and K'Sante; jungler Tian picked Nocturne in two of three. On the surface that reads as a shallow pool. A stricter reading says the opposite. In BO1 there is no pool-exhaustion penalty, so repeating the most stable priority pick is rational, not lazy. TES's behaviour confirms the format's incentive structure — evidence for the thesis, neutral about the team. There is a fundamental methodological problem here: BO1 cannot measure pool depth. A player with twelve viable champions and a player with two behave identically in a best-of-one. A three-game sample therefore supports two equally valid explanations — one, the team's pool is narrow; two, Ambessa and K'Sante are simply the two dominant top laners of this patch, making this meta concentration rather than conservatism. Without patch version, win rates or pick/ban baselines, the two cannot be separated. Any conclusion about TES is directional at best. Which is why calling the current hybrid an oversight is also too simple. A more defensible reading is a deliberate two-layer design. The BO1 layer — rounds one to three — delivers high variance, high drama and zero pool tax; it works as a handicap system that keeps minor-region and narrow-pool teams alive and generates upsets. The BO3 layer — rounds four and five — activates the pool tax and rewards depth exactly when elimination is at stake. Under that reading the format is not incoherent; it is trying to serve several objectives at once. The failure is narrower: the diversity Riot publicly promised is the one thing this design shows least. It is also worth remembering that BO1 is not a Riot invention. CS2 Majors run BO1 for non-decider Swiss matches and BO3 for advancement and elimination. DOTA2's TI uses BO2 group round-robin, avoiding BO1 entirely. Valorant's VCT uses BO3 groups and BO5 finals. BO1 appears wherever match volume is highest and broadcast windows are tightest — a throughput solution, not a competitive philosophy. That weakens the claim that Riot alone is getting it wrong; it does remain an outlier within Riot's own system, since its other events have largely abandoned BO1. Economically, the real constraint is broadcast hours, not competitive philosophy. Converting rounds one to three to BO3 would lift total Swiss games from roughly 56–72 to roughly 80–120 — a 40 to 65 percent increase in live content — which means extra competition days, arena rental, production crew, travel and accommodation, all absorbed directly by a first-party publisher. That is why the BO1 compromise survives. The commercial contradiction is that Riot pays the cost of Fearless Draft across the whole Swiss stage — pool pressure on players, preparation burden on coaching staffs — while collecting its product benefit of champion diversity, highlight clips and differentiation from rival titles in only about 40 percent of matches. At the governance level this is not a rule violation but a policy coherence gap. Riot publicly committed to a diversity objective, and its own format undercuts that objective in 60 percent of matches. In publisher-governed esports there is no independent arbitration body, which makes exactly this kind of mismatch the main source of credibility erosion. The counter-weight is that a three-game sample on one team is an equally serious methodological weakness for anyone drawing big conclusions. The rule's definition is also under-specified. 'A champion already used cannot reappear within the same match pair' leaves at least three things unsettled: do banned champions carry over? does a champion played in two roles count once or twice? does a Swiss rematch inherit the pool? The answers could nearly double the size of the pool tax, and the first real dispute will surface in a knockout BO5, where the stakes are highest. One risk is under-priced: the cold-start cliff. A team that goes 2-0 or 0-2 in BO1s walks straight into a BO3 decider that is its first best-of-three of the tournament, with the pool tax switching on abruptly and no acclimatisation. That damages not just the team but the product: the one-mistake elimination of a marquee side in a BO1 hurts viewership for later rounds, which runs against Riot's own commercial interest. The most likely outcome is therefore incremental reform, not a redesign. Riot's track record is one of repeated, small format adjustments. The most practical route keeps round one as a BO1 spectacle slot and starts BO3 from round two or three — activating the mechanism earlier while keeping broadcast hours in check. An all-BO3 Swiss is the cleanest fix technically, but calendar and production costs make it the least likely in the near term. What to watch: first, champion-uniqueness rates in BO1 versus BO3 rounds across the whole Swiss stage — if BO1 is materially lower, the argument gains real data. Second, how many champions 369 uses once the format switches; three or more in BO3 would falsify the shallow-pool claim outright. Third, whether Riot clarifies whether bans carry over. Fourth, the next season's format announcement. Fifth, viewership of BO1 versus BO3 rounds. Until then, the essential caution is not to treat one team's three games as a trend. This is sports-news analysis based on public information, not betting advice; the TES champion data is a three-game sample pending verification.