iTero, GIANTX and the Undrawn Boundary of AI Coaching in Esports

Inside the roughly 12 minutes between game 1 and game 2 of a best-of-five, there is something no stat sheet ever records: how fast a coaching staff can rebuild a draft plan. I have sat through enough green rooms and enough post-match reviews to know that the biggest series are rarely decided by in-game execution. They are decided in that silent gap, when one team reopens its data and the other team simply tries to remember by feel. That is why the conversation around Jack Williams, the iTero tool, and the GIANTX organisation deserves to be pulled out of the news briefs. When an AI coaching platform signs an exclusive deal with an organisation competing in a regional league, the question is no longer whether the tool works. The larger question is this: a preparation advantage that can be bought, locked exclusively, and denied to everyone else in the same league, belongs to which category? This is the grey zone the whole industry is standing in. Fans are used to two debate frames: one says AI tools professionalise esports, the other says AI cheats. Both frames come with ready-made headlines. But the third frame, unequal access inside a closed league, is rarely discussed, even though it is the one that determines the real value of every deal like this. Context: in a closed league, a temporary advantage becomes a permanent one. In an open-circuit system, an advantage erodes on its own. The winner gets scouted, copied, and imitated the following season. But in a franchised league, where member teams are not relegated, structural advantage is never competed away. It survives across seasons. If a single team in a league holds the right to use a tool nobody else can touch, you are not talking about a commercial contract. You are talking about the league operator accidentally permitting an inequality in preparation. This is exactly what I have always argued against: home advantage is an illusion, and exclusive tooling advantage is the same illusion. People pour money into flashy variables to hide the real variable sitting inside the analysis room. Core analysis: the value of AI is not fixed; it inverts with patch cadence. This is the part almost nobody counts. An AI coaching tool has no absolute value. Its value depends on how fast the underlying game changes. In titles with rare, large, disruptive updates, historical data keeps its validity for a long stretch. Machine-learning tools gain an edge in modelling depth. In titles that patch quickly and continuously, the half-life of any learned pattern collapses. At that point, the value of AI stops being about solving the meta and starts being about detecting the meta delta faster than opponents. That is a speed advantage, not a knowledge advantage. Two completely different mechanisms. A product marketed identically across both is a red flag. Because if AI only accelerates information processing, it is really selling time, not intelligence. Time can be bought. Coaching intelligence cannot. This is where I break from the crowd. The industry is debating whether AI is cheating. Wrong question. Cheating is only the crude version of a bigger problem: access. Every argument about AI in esports, once you peel off the moral layer, circles back to a distribution question. And one thing should be said plainly: any performance claim about an AI tool is meaningless without a disclosed sample size and evaluation method. I no longer use the words locked or confirmed before a contract is signed, and I will not accept performance numbers with no method behind them either. If the ink is not dry, do not call it a blockbuster. Contrarian angle: the real barrier is not publisher rules, it is exclusivity contracts. Everyone is waiting for publishers to ban AI. That is the wrong place to wait. Real-time in-game assistance has been clearly banned in every major title for years, so there is nothing left to debate. The real grey zone sits in the window between games, the break inside a best-of-three or best-of-five, where no rule was ever written for a tool that can re-read an entire dataset and propose adjustments within minutes. That is why exclusivity matters more than cheating as a word. If publishers do not ban it, the only thing protecting fairness is league regulation. And if the league does not mandate equal access, exclusivity quietly becomes the law. I could be wrong here. If the tool only processes publicly downloadable data, then exclusivity is just exclusivity of deployment speed, not exclusivity of data. In that case the competitive impact is far smaller than the fear suggests. Which exception could falsify my numbers? The answer lies in whether that contract locks the organisation's private data. Nobody discloses that. And silence, in this case, is not neutrality. Takeaway. People laughed at my predictions, but nobody laughed at how I recounted every number. Here is the specific thing I will track: over the next two seasons, whether the league forces every analytics tool to open equal access to all member teams. If it does, we know the operator treats preparation as part of competition. If it does not, we know it treats preparation as a permitted investment. To fans, this sounds distant. But remember: a good hot take is not about daring to be wrong, it is about daring to be right before the whole world. And the real question is not whether AI is changing esports. The question is who gets to pay for it first.