Does MSI Predict Worlds? Six Titles and a Sample That Is Still Too Small

On May 21, 2026, in London, JDG lifted the MSI trophy after a five-game series against BLG. On November 19 of the same year, at Gocheok Sky Dome in Seoul, T1 beat Weibo to win Worlds. Two different champions, two completely different settings. But only days after the final in Seoul, a tidier version of the story began spreading across analytical forums: MSI is no longer a meaningless tournament, it has become an accurate test for the year-end crown. The evidence offered was compelling: the six most recent MSI titles all belong to JDG, BLG, Gen.G and T1, exactly the four organizations that have shared almost every international trophy in the same window.

That number is not wrong. What it is used to prove is.

For most of this discipline's history, MSI was treated as a trap. In 2026, RNG won MSI and exited Worlds in the quarterfinals. In 2026, G2 won MSI and lost the Worlds final. In 2026 and 2026, RNG repeated the same script, winning MSI and leaving Worlds in the quarterfinals. Even JDG, the strongest team of 2026, won MSI and then fell to T1 in the Worlds semifinal. In Seoul, where I live and work, people call it the MSI curse, and for years it was a more reliable statistical rule than any expert opinion.

In 2026, Riot Games changed the format. MSI expanded its field, added a play-in stage and, most importantly, a lower bracket. To win, a team had to take more best-of-five series, face more opponents from more regions, and adapt to multiple playstyles within a single event. Structurally, MSI moved closer to the shape of Worlds. That part of the story is true. The rest is where I want to stop.

Break the six most recent MSI titles down into raw data and the picture changes. The 2026, 2026 and 2026 seasons contained three MSI events and three Worlds events, six major international titles in total. They are not evenly distributed. Gen.G won MSI twice but never touched the Worlds trophy in that window. T1 did the opposite: three straight Worlds titles without a single MSI crown. JDG has one MSI. BLG reached multiple international finals without winning any title. At the individual level the gap is even clearer. Chovy holds two MSI titles in the same period but has never won Worlds, while the T1 core built around Faker lifted the Worlds trophy three seasons in a row.

The claim that the MSI champion also wins Worlds therefore does not hold if read as a causal chain. What does hold is a different, narrower and far less glamorous claim: the pool of organizations capable of winning any international event has shrunk to four names. Those four names have taken every final. When the candidate pool is only four, the probability that two tournaments in the same year pick their champion from the same group is very high, as a mechanical consequence rather than as proof of a format's predictive power.

Core insight: the overlap between MSI and Worlds over the past three years is a consequence of the championship candidate pool shrinking to four organizations, not evidence that MSI has become a reliable forecasting tool.

I have used this line of reasoning before. In 2026, when the K-League returned without crowds, I collected data from the first 42 matches and found the home win rate had dropped from roughly 40 percent to 25 percent. The conclusion I drew was not that home advantage had vanished, but that home advantage was never an independent variable. The MSI case is the same. The overlap may not come from MSI becoming a better test, but from the number of organizations genuinely capable of winning dropping to a historic low.

Money is the first cause. These four organizations operate on the two largest payrolls in the two largest regions. During the transfer window, they are the only teams able to fund an all-star roster without selling a cornerstone. When a market has only four real buyers, the results on stage also have only four real sellers.

Another cause is roster continuity. BLG and Gen.G have kept stable cores across multiple seasons, something most teams elsewhere cannot manage. Transfers are a game of reading a manager's ego, not a game of buying and selling. Every transfer window, dozens of teams swap three or four positions and then enter international play with far lower cohesion. That stability never appears in a power ranking, but it lives inside the results.

The most undervalued factor is the patch environment. The gap between MSI and Worlds stretches four to five months, enough for two or three major meta shifts. In all three recent seasons, Worlds was played on versions substantially different from the one used at MSI. No win-rate, pick-ban or game-length data has been published showing a systematic meta shift between the two events, and the absence of that data is itself a problem. A claim not yet proven by numbers should not serve as the foundation for a prediction.

There is a subtle point that supporters of the new claim usually skip: MSI identifies contenders far better than it identifies champions. A team that goes deep at MSI is almost certainly a team good enough for Worlds, but the ranking order between the two events does not correlate. That makes second or third place at MSI potentially more interesting data than the title itself, especially since a lower bracket makes placement depend on bracket draw and timing.

It is worth noting that MSI's new format was never announced with any commitment to synchronize patches with Worlds. The fact that MSI and Worlds share a broadly similar champion pool is a byproduct of the current balance cycle, not of deliberate design. If the update cadence shifts in 2026, the entire correlation under discussion resets from zero, and that is the largest risk for anyone building a predictive model on three years of data.

Regionally, all six titles belong to the LPL and the LCK. No team outside those two regions reached a final in that window, a far higher concentration than in 2026-2026, when European and North American teams still appeared in the last match. The smallest detail in the room usually says the largest thing, and here the detail is that the number of teams capable of winning a best-of-five against the LCK or LPL is falling, not rising.

Behind that concentration sits a problem rarely discussed. When four organizations absorb nearly all resources and titles, the rest of the world has little incentive to invest in grassroots coaches and youth development, things that take five to seven years to pay off. Academies bearing the names of former stars keep appearing, and most operate more like a media channel than a training centre. The result is that the gap is not closing but being recreated every season, and that gap is precisely what produces the illusion that MSI has predictive power.

The competitive calendar is another forgotten variable. Thin rosters enter the final stretch with higher injury risk, while deep rosters, which is what those four organizations have, absorb less damage. Load management and rotation are usually framed as a story about professionalism, but in practice they work as a familiar filter: the team with money keeps its people.

So where might I be wrong?

Three possibilities. The first is that the expanded format genuinely improved signal quality, and those four organizations are simply the fastest to adapt to a better test. The second is that the concentration of power has become the new stable state of the discipline, in which case the phrase MSI predicts Worlds will be true, but true for a banal reason rather than because MSI has magic. The third is that I am misreading the data: a sample of six titles across three years is far too small to reject anything, including the hypothesis I have just built. If you are right before your time, you are called a madman. If you are right after, you are called a genius.

To stop the debate from running forever, I set three specific verification criteria. If an organization outside the LPL and LCK wins MSI 2026, the candidate pool has widened again and every conclusion drawn from the past three years loses its foundation. If the MSI 2026 champion is eliminated before the semifinals of Worlds 2026, the forecasting model collapses in the most literal sense, just as it did for RNG in 2026 and 2026. And if over the next three seasons the international semifinal slots still revolve around the same four names, then we have gained no evidence at all, only three more years of the same phenomenon.

People say I object just to be noticed; I am simply looking one step ahead. I do not listen to the crowd; I read the draft.

When MSI 2026 begins, the thing worth tracking is not who lifts the trophy. The thing worth tracking is who reaches the semifinals, and whether any name falls outside those four organizations. If none does, the past three years have proven nothing. If one does, we will finally have a data point with real weight. And if the MSI 2026 champion goes on to win Worlds, I will be the first to rewrite this piece, and that time I will not need to persuade anyone with an argument at all.