Chovy and Rookie: Seven Years for a Rematch, and the Data Gap in Between

October 2026, Madrid. I sat in front of a screen with a notebook and four ruled lines. Invictus Gaming against Griffin in the Worlds quarterfinal. Rookie on mid for Invictus Gaming. Chovy on mid for Griffin. Four games. 3-1 to Invictus Gaming.

Four games is the entire number of times those two have stood in the same official match at an international event. Seven years, counting to today.

I wrote those four lines down because at the time I was teaching myself to chart a mid laner's movement: wave state, roam timing, distance to the jungler. I did not know what I was recording for. I only knew that four games was too small a sample to say anything certain.

Seven years later, there is an article about that rematch. Long, coherent, emotional. Of its 21 information points, 12 are marked as fact, and not one of them carries a verifiable source. Not a line about the patch. Not a line about champion pools. Not a line about mid-lane pick priority.

That is why I reopened the old notebook.

Chovy — Jeong Ji-hoon — in 2026 was the youngest mid laner in the group people called the Griffin generation. Rookie — Song Eui-jin — had already won Worlds 2026 with Invictus Gaming and was the only Korean in the starting lineup of a Chinese team that won the world championship. In Madrid, Rookie won the mid lane in a way the scoreboard barely showed. Invictus Gaming then lost 1-3 to FunPlus Phoenix in the semifinal.

Since then the two have taken different roads and have not met in an official international match. Chovy has stacked LCK titles with Gen.G, has not won Worlds, and his best finish is a semifinal. Rookie left Invictus Gaming, passed through a few teams, and came back.

A rematch after seven years is rare. In mid lane, where the average professional career is far shorter than a footballer's, seven years is close to a generation. The interval itself generates the pull.

Pull is not evidence.

Seven years is the gap between two meetings. It measures nothing about the quality of the second one. In sport, people use a time gap as a proxy for rarity, then use rarity as a proxy for value. That chain breaks at the second link. A rare match can be a great match. It can also be a poor match remembered only because there had not been one for a long time.

I work by one rule: when an esports article never mentions the patch, I read it as an article about the schedule, not about the match. The patch is the one variable neither player controls, and the variable that decides who holds the edge. Data never lies, but whoever defines it can.

Among those 12 facts there is one worth pausing on: the article states that Chovy needs more convincing performances on the international stage. I do not dispute the conclusion. I dispute the delivery. A claim about international performance with no win rate, no game count, no opponent attached is an opinion dressed as a measurement. That kind of sentence travels fastest and verifies worst.

Riot's established practice is to lock a dedicated tournament patch, published a few weeks before the event begins. The consequence: by the knockout stage the meta is frozen and no longer shifts in response to what teams do. The adaptation race happens before the tournament, in the practice room, not on stage.

Which means most of the story about who adapts better was already written before the first game. And it was written by people who never appear on screen: the analyst staff.

Since 2026, Worlds opens with a Swiss stage and moves into BO5 knockout. I have watched enough Swiss rounds to know one thing: no team gets through that stage on a single composition. You need at least three distinct draft directions, and all three have to be meta-legal. A narrow champion pool does not die in game one. It dies in the fourth draft.

A solved meta has a consequence nobody talks about: it turns a wide champion pool into a defensive tool rather than an offensive weapon. When every team knows the same ten priority picks, the player with range does not use it to impose a shape on the game; he uses it so that a shape cannot be imposed on him. Its value sits in the third and fourth draft, when the opponent has run out of directions and you still have a legal answer. For a mid laner, that is the whole difference between advancing and stopping.

Mid lane makes this harsher. The mid laner is the first player forced to widen his pool, because mid is where teams concentrate priority picks and where counter-picking is easiest.

Here is the part I want to state plainly: I cannot say who holds the edge, because the source article provides not a single line about either player's current champion pool. Any conclusion I offered on that point would be fabrication with decoration.

What I can do is build two conditional scenarios.

Scenario one: the locked patch favours lane-dominant, resource-rich mid laners capable of solo carrying. In that world the individual skill gap is amplified and the edge tilts toward whoever is at peak form. Scenario two: the locked patch favours supportive, roaming mid laners who concede resources to the bot lane. In that world the skill gap compresses and the match moves closer to a coin flip.

Neither scenario is a forecast. They are frames for reading the result after it happens, and for knowing where I was wrong if I was wrong.

At this point I have to pull in another metric, one I learned from a failure.

In 2026 I built a model for the Euros and predicted Italy would go out in the quarterfinals, because they generated only 1.2 expected goals per match. Italy won the tournament. Rewatching the tape, I found the thing I had never measured: the average distance between their two centre-backs was 21.4 metres, the smallest in the tournament. They did not create more chances than their opponents. They stopped chances from forming before they became shots.

League of Legends has an equivalent metric nobody has named properly: the average distance between the mid laner and the jungler over the first 10 minutes, sampled per second, split by wave state. It does not tell you who wins. It tells you who controls the tempo of the map.

I tried to build that metric by hand for the four Madrid games, from public footage. The result: in the two games Invictus Gaming won, Rookie kept a markedly shorter average distance to his jungler than in the other two, and his roam timings lined up with the third wave more often. Four games. Not enough to conclude anything. Enough to show that anyone writing a real analysis of this rematch has to start there, not from the seven-year story.

I also have to state the limits of that measurement, because I have published a wrong model before. Public footage gives no continuous coordinates, so I had to sample frames and interpolate, with an error of at least half a second each time. Distance to the jungler is also shaped by where the jungler is, and jungler position depends on which lane is pushing — an endogenous variable, not an independent one. And four games is not a sample. Four games is an anecdote with numbers attached.

Back to Madrid one more time. Invictus Gaming won 3-1, but the game they lost was the one Griffin controlled from the eighth minute. I remember it because it broke how I was reading matches then: I was looking for who played better, while what actually changed was who controlled when the fights happened.

This is where I have to argue against myself.

A rematch after seven years is compelling because it assumes the two players are still the two players they were. That assumption is structurally wrong. Seven years in mid lane is seven years of changing champion pools, changing reflexes, changing map reading, changing role inside the team. Comparing two versions of the same person is a comparison with an uncontrolled variable.

There is another layer. The source article itself concedes that the result may not fully determine either player's tournament run. That is an honest sentence, and it quietly demolishes the article's own premise. If the result decides nothing, why wait seven years?

The answer sits outside competitive data. It sits in the fact that audiences need a story with a timestamp. Seven years is a good timestamp. It is easy to tell, easy to remember, easy to share. It does not have to be correct.

I understand that pressure. At Northampton, when I wrote a 40-page report on PPDA, nobody read it. When I compressed it into one line — the team is defending actively, not attacking chaotically — the manager read it. At Northampton we had no technology; we had patience and one spreadsheet. But the larger lesson was this: a finding only has value when it is told in a way someone can act on.

So I do not object to the seven-year story. I object to telling it without conditions. A wrong yardstick is more dangerous than no yardstick at all.

There is one more variable the article leaves blank, and it concerns both players: career length. A professional League of Legends career is far shorter than a footballer's. But the academy system and post-retirement support in League of Legends are close to zero compared with football. Those seven years include years without competition, years of practice without a tournament, years in which body and reflex are consumed inside an industry that has no insurance mechanism for that asset.

That is the thing worth measuring. Nobody counts it.

I do not trust intuition; I trust data — and data is what taught me to trust nobody, including a good article.

Next cycle I will watch three things.

The champion list locked for the tournament patch, published before the event begins. That is the real power map of the tournament, and it exists before any prediction.

Mid-lane pick priority in both teams' first Swiss game. That is the first testimony about which team understands the meta better, and it cannot be faked.

The average distance between mid laner and jungler over the first 10 minutes. I will measure it myself, by hand, and I will publish the error bars.

Every number is a story waiting to be verified. Seven years of waiting is a good story. But if by the tenth minute of game one none of us knows how far the mid laner is standing from his jungler, we have not watched that match. We have only been rewatching our memory of it.