When One Match Is Not Enough to Conclude: Croatia 2026, the Empty Stadiums of 2026, and the Morocco Matrix

On 11 July 2026, at Luzhniki, I was 23, a master's student writing a tactics blog. Before the Croatia–England semi-final I published a confident prediction: Croatia would press high through the Modrić–Rakitić–Brozović trio, win the ball in the opponent's half and control the game. The match went the other way. Croatia pushed up for exactly 18 minutes, then dropped deep into their own half and let England hold 57% of possession. The final score after extra time was 2-1 to Croatia, built on the space behind England's back line. I got the winner right and the mechanism completely wrong.

That night I wrote a 1,200-word self-critique. I admitted I had judged people instead of space: I looked at the names of three Croatian midfielders and inferred a style, rather than looking at how they occupied the squares of the pitch. The piece happened to be read by an editor at a sports-data company in Seoul. He sent me a short email asking whether I wanted to try working with positional data. Eight years later I am still in Seoul, still an analyst, and still carrying the biggest lesson of that night: one match, however dramatic, is not enough to conclude anything.

Modern football analysis runs on a paradox. Fans want answers the moment the whistle blows. Media platforms reward speed, and the slower analyst gets pushed down by the algorithm. I understand that pressure better than most, because I write for the Korean market from Seoul, where every round ends with hundreds of headlines fighting over one angle. But I am also someone who has paid for a rushed conclusion, so I forced myself to build a process that runs against the crowd.

I joined the sports department of Belgrade Television in 2026, very young, and those years taught me that sports news is not just retelling a scoreline. I took part in publishing works such as “Tip Off”, “Chasing the Game” and “The Pine Tar Game”, and each project taught me the same thing: readers forgive slowness, but they do not forgive fabrication.

My process has three rules. First, every analysis must carry at least three data points on space, distance or line spacing; without those three figures I am not allowed to write a single concluding sentence. Second, I stopped writing emotionally about player reputations and instead used coordinate maps for every notable phase, to see who moved, where they moved and which gap they left behind. Third, and most importantly, I must state clearly what I am missing before I state what I know.

The third rule sounds so modest it is almost boring, but it is the boundary between analysis and invention. At work I constantly receive incomplete datasets: a match with event data but no positional data, a player with only 300 minutes all season, a team three games into a new manager. Faced with such data there are two ways to behave. The first is to fill the gap with intuition and present it as a firm conclusion. The second is to say plainly that there is not enough to conclude. I choose the second, and I am grateful my profession allows it.

To see why saying “not enough” is so hard, look at three times I stood before an empty or near-empty dataset.

The first was that Croatia–England match in 2026. My dataset then consisted only of what I remembered and what I had read in newspapers. I had no figures on the distance between Croatia's lines, no heat map showing how deep they dropped, no real pressing minutes. Yet I still produced a very specific tactical conclusion. My mistake was not predicting wrongly; my mistake was predicting at all when I lacked the data to do so. When Croatia came back, I understood that football is not mathematics, it is ethics. One match can teach you about the courage of a group, but it cannot teach you about a system if you have only one sample.

The second was the summer of 2026, when I was 25 and working as an analyst at a sports-data company. Covid-19 forced K League 1 to play without spectators, and I was handed a paradox. The league's average home advantage fell from 1.48 points per match to 1.12 points per match, and the gap held steady across roughly 200 matches. At first I set the result aside, because it broke every precedent I had learned about home advantage. But I could not set it aside forever. It took me three weeks to re-run multiple models, cross-check week by week and team by team, gradually remove the pandemic factor and the confounders, and only then publish an internal report. That was the first time I saw a tactical change come not from a coach but from absent spectators. Football without crowds in 2026: every tactic was still correct, yet no tactic still meant anything.

The third was the 2026 World Cup, when I was 27 and assigned to track Morocco's entire run. This time I had complete data, and I spent four weeks rewatching every match. I counted how often Hakimi and Mazraoui tucked inside, recorded the average distance between the two central midfielders at 12.4 metres, and noticed that the open space in front of the box was always screened by an inverted triangle. Morocco conceded only one own goal in the group stage; the other two goals came in the semi-final against France. I published “Morocco's Defensive Matrix: Space Under Occupation”, and it was shared more than 2,000 times in the Asian tactics community. The Morocco matrix was not there to block the ball, it was there to strangle the opponent's time. The difference between the third case and the first was not my talent; it was that I finally had enough data to conclude.

Those three cases taught me something it took years to put into words. The value of an analyst is not in reaching conclusions fast, but in identifying the exact moment the data is sufficient to conclude, and in having the courage to say “not enough” while it is not. A conclusion built on an empty dataset is not analysis; it is a fabricated product presented in a nicer font.

Small samples are the enemy of every tactical conclusion. At a major tournament each team plays only three group games, and those three games usually fall inside ten days. With a sample that small, everything becomes a signal: a goal from a set piece, a red card, a long-range strike into the top corner. I have seen teams praised as having a “new identity” after one win, then read completely differently after the next match. I set myself a threshold: I do not declare that a system has changed without roughly 900 minutes, about ten matches, under the same manager. Below that threshold I am allowed to describe, not to conclude.

This is especially true in the knockout rounds of national-team tournaments. A lowly-rated side that reaches a final is usually told as a story of a successful system. But look closely, and most of those runs come from a favourable bracket and one explosive night. That is the luck of the draw plus one match at peak form, not evidence of a repeatable model. I do not say this to flatten the emotion of fans; I say it because if we call luck a system, we learn the wrong lesson and carry it into the next match.

Even in areas that seem purely emotional, such as how referees treat big clubs and small clubs, I try to keep data discipline. I believe the difference in how fouls are called is not a conspiracy, but a consequence of real crowd and media pressure. Yet to prove it we need thousands of consistently coded incidents, not a few controversial slow-motion replays. When the data is not there, the most honest thing to say is: I suspect it, but I have not proven it.

The same principle applies to the transfer market, where I have witnessed many rushed conclusions. The transfer market is a market of regret: whoever can wait wins, whoever rushes pays. Neymar's move from Barcelona to Paris Saint-Germain in August 2026, at a record fee of 222 million euros according to PSG's official announcement, is an example of how one deal can reshape an entire price level within weeks. But to judge whether a deal succeeded or failed you need at least two seasons of data, not one unveiling.

Based on my experience watching matches across the Vietnamese and Korean football scenes, I have learned that fans are not afraid of caution. What they fear is fake caution, the analyst who never dares to say anything specific. So I distinguish very clearly between two states: not enough data, and enough data but no willingness to conclude.

All of this leads to a concept I learned from the data work itself: the null result. When a model runs and returns no signal, that is still a result, and sometimes it is the most important one. It says that our question was not framed correctly, or the data was insufficient, or the phenomenon we were looking for does not exist. To an analyst, a null result is not a failure; it is a reminder that honesty is worth more than appeal.

But I must say plainly something many colleagues avoid. There are two very different kinds of “not enough data”. The first is genuine humility: the data is missing, and we honestly say we cannot conclude. The second is intellectual cowardice: the data is actually sufficient, but we still plead “not enough” to avoid making a judgement and avoid responsibility if we are wrong. I have fallen into the second kind, and it is more dangerous than it looks, because it dresses avoidance in a moral coat.

The biggest blind spot in this profession is not a lack of data. The biggest blind spot is a system that rewards decisive conclusions and punishes silence. When the whole industry pays for people who speak firmly, the honest analyst is pushed into choosing between truth and attention. I choose truth, but I am aware that every time I say “not enough”, I face a new temptation: to turn caution into an escape route.

So I set myself a balancing rule. I believe in structure, but structure exists to collapse; a good analyst is one who predicts the point of collapse accurately. Once the data is sufficient, I must conclude, and the conclusion must be specific enough to be wrong. I state the counter-evidence in two sentences, then lock my position with one index, one mechanism and one time frame. If I am not willing to bet on what I have just said, then I have not really analysed; I have only described.

Going into the next match, readers can ask themselves one simple thing whenever an expert declares something after a single game: how many samples, how many metres of distance, and how many minutes of football does he have to say it. The answer tells you whether this is analysis or just an impression repackaged. And if someone, including me, says there is not enough data to conclude, that is not the weakness of the analyst. It is the only position from which an honest conclusion can begin.