MSI and Worlds: Six in a Row and the Eighteen-Month Statistical Trap

In my analysis room, I keep a handwritten notebook. Its first page has exactly one line, written in blue ink: "May 21, 2026 – JDG lift the MSI London trophy, 3-2 against BLG." Below that line is an empty box. I have deliberately left it blank for two and a half years. Not out of laziness. Because I am still not sure what should go in it. Every time an MSI cycle closes, the fans ask the same question: "Does the MSI champion go on to win Worlds?" That is the easy question. The hard question lies elsewhere: are we seeing a rule, or are we seeing a convenient arrangement of memory? Throughout this regular season, as I tracked every match and noted every key timestamp, I kept finding that the answer is not on the standings board. It is in the place nobody wants to look at directly: the size of the data sample. I first noticed this on a summer evening. On screen was the MSI bracket, slowly closing. I counted the teams that had won MSI and then later won Worlds. The result came out like a round of applause: six out of six. JDG. BLG. Gen.G. T1. Those names repeated like a refrain. But when I closed that table and looked out the window, I did not feel reassured. I felt that something was off. To understand why, we need to step back a little into MSI’s history. The mid-season event was born with a clear mission: bring the world’s top regions together in the middle of the year, create a checkpoint of form before the season pours toward Worlds. For many early years, MSI worked like a light tap. People watched it for entertainment, for form, to see which region looked sharper. Almost nobody treated it as a forecast for the year-end title. Then the organizers changed the format. MSI was expanded. More teams, more matches, a longer road to the trophy. On the surface this sounds like a logistics story. In reality, it changed the nature of the tournament. An expanded MSI was no longer a weekend excursion. It became an event that demanded continuous adaptation, roster depth, fast meta reading, and fast error correction. In other words, it began to demand exactly the qualities Worlds had always demanded. That is why I think the recent overlap is not entirely random. But "not entirely random" is different from "a rule". Between those two things lies a gap, and that gap is the place I want to stand in while writing this piece. I rewatched the footage of each recent MSI final. Not to find highlight plays. I watched to count. To count the minutes the winning team controlled the flow of the game. To count how often they opened fights on purpose. To count how often they backed off at the right moment to trade a small objective for a bigger one. To count the silences during respawn timers, when the lineup stands still and the head keeps running. What I found caught my attention more than the results table itself. The recent MSI champions did not win with a single style. JDG won with clear structure, almost technocratic. BLG won with draft trickery and the ability to shift tempo in an instant. Gen.G won with macro discipline and a patience that sometimes made viewers anxious. T1 won with something else entirely: calculated explosion, where a single personal play can overturn an entire game. Four squads, four languages. Yet all of them could say the same sentence in the language of the map: "We know where this game is going before you do." That is the thing an expanded MSI began to measure. And that is the thing Worlds has always worshipped. Now let us return to the number that kept me up at night: six out of six. Over the last three years, each MSI produced teams that later stood at the top of Worlds. JDG in 2026. BLG in 2026. Gen.G in 2026. BLG in 2026. Gen.G in 2026. T1 in 2026. Six MSI titles, six conversions into year-end glory. Read that far and the first reflex of most fans is to nod. "MSI is a good indicator now." And if I were a writer who only wanted to please the majority, I would stop right there. But I am not that kind of writer. I am the one who sits in an empty host booth logging 47 timestamps of waiting during a tournament with no crowd. I am the one who rewinds a tape four times just to note where a support placed a ward in the twenty-seventh minute. I know that behind a beautiful number there is always an ugly question waiting. And the ugly question here is: how big is our sample? Let us count carefully. Three years. One MSI cycle per year, though the expanded format allows more than one team to be crowned within a year in some aggregated data systems, so the final figure I hold is six. Six. In statistics, six data points are not a rule. They are an anecdote. They are a story told at a café table, not a law written in a textbook. I want to say this slowly, because I know it is hard to hear. A streak of six out of six is far less likely than flipping a coin six times and getting six heads. But low probability does not mean impossible. And in a discipline where only one team wins Worlds each year, six consecutive results lining up still falls inside what I call "noteworthy but not yet believable". The meta does not die; it molts into another poem. The same is true of statistics. A small sample does not disappear. It merely molts into a belief that people no longer verify. Why am I so strict about this? Because I have already seen a belief shattered in front of me. In 2026, I sat in Kazan, watching a team once held up as the model of possession football get eliminated in the group stage. The crowd roared; I bent down and opened my laptop. I rewatched seven of their matches and saw something identical to what I had once seen on the map of a video game: a possession-dominant style with no finishing damage. An outdated build. When Germany collapsed, I understood that an ideology also has an expiration date. That lesson has followed me throughout my career. It keeps me from ever reading a single match as an isolated event. Everything is part of a longer cycle: the cycle of the ruleset, of conditioning, of team psychology, of how fans remember a victory. And a cycle, because it is long, always demands more data points than a few years can supply. In the context of the LPL–LCK contest — the two regions that produced all six MSI titles in the period under review — a small sample becomes even more dangerous. When only two regions split every title between them, it is very easy to mistake a phenomenon of concentrated power for a rule of the game. You see China and Korea trading the top spot, and your brain automatically encodes it as "the summit is in one of these two places". But that is not correct. It only means that, during the window we observed, no other region could touch the trophy. People think they are reading the match, when in fact the match is reading them. What does the match read in us? It reads that we prefer a story with a clear plot over a series of random events. It reads that we are willing to turn a short-term trend into a long-term truth, just to feel that this world is predictable. And it reads that we, the followers of esports, are so used to believing in metas that we forget a meta also has a lifespan. I have spent many nights looking back at the runners-up and third-place teams of recent MSIs. That is the part where I think the forecasting problem becomes far more interesting than simply looking at champions. A runner-up at MSI is not a loser. They are a team that reached the final, beat many strong opponents, and proved their ability to adapt against multiple styles. In the old format, that achievement could be overshadowed by someone else’s title. In the expanded format, it becomes a signal of its own. I believe this is something many analysts overlook. They focus on the question "who won MSI