The Empty Report: When Sports Analysis Contains Not a Single Line of Data

In July 2026, in Saint Petersburg, I sat in the media area of the World Cup semi-final between France and Belgium. Above me, a large screen displayed the value of the broadcasting rights American networks had paid to air the match. Below, Russian supporters bought tickets in roubles at a price that was a fraction of the figure overhead. I recorded both facts on the same page of my notebook. Three weeks later I built a private cost-benefit model to quantify the gap, then deleted it. The dataset was not large enough to guarantee reliability, and in my line of work a wrong conclusion does more damage than no conclusion at all.

Seven years later, I received a long document. It had every section, every table, every bold heading. Opening it, every cell said the same thing: insufficient information to assess. No tournament name. No team name. Not a single metric. The author had refused to invent. In an industry where everyone is trying to appear more informed than they are, it was the most honest document I had read in years.

But it also exposed a problem larger than itself. An empty analysis is a paradox of the whole industry, because it only exists when someone commissions it. And people commission it because in modern sport, "analysis" has become a commodity traded more frequently than players.

The supply chain of a belief

The sports industry runs on an information supply chain with five clear links. At the source sit game publishers and tournament organisers, producing raw data: balance patches, schedules, transfer rules, match statistics. In the middle sit clubs, agents and analytics departments, where raw data becomes decisions. Downstream sit media and fans, where decisions are converted back into stories. Each link sells the next one something collectively called analysis. What is actually sold is mostly conclusions, not evidence.

In esports, a meta report can be written from twelve matches in one regional event, then cited as gospel for an entire international season. In football, a scouting dossier can rest on three four-minute highlight reels, with no real minutes played, no opponent context, no injury data. Both pass through the same gate: the reader has no way to verify, and the writer has no incentive to self-verify.

The cost is not that the information is wrong. It is that the information cannot be contested. A conclusion with data can be refuted, and that very possibility is what gives it value. A conclusion without data can only be believed or resented, and neither reaction improves anyone's decision quality.

The transfer market is where this mechanism is most visible, because it is where belief is priced into real money. A player rumoured for three weeks can see his asking price rise twenty percent without playing a single extra minute. A club can spend a large sum on a viral clip, then discover the clip was cut from two matches against the weakest opponents of the season. Every transfer bubble begins with a beautiful story and ends with a balance sheet.

What stands out is that the industry does not lack data. Public statistics platforms now track thousands of matches per season, in both football and esports. Tournaments release positional data, passing data, per-minute resource data. The problem is that this enormous volume has never been converted into questions. People collect because collecting is cheap, and analyse because analysing is cheaper still.

The data is not missing. The questions are.

In 2026, during the European Championship, I built a small database with a single criterion: players under twenty-one, fewer than five hundred league minutes, but with a high pressing index. The criterion sounds narrow, and it was narrow. I narrowed it deliberately, because I had seen too many scouting reports built from thousands of players that ended by picking the most famous name on the list.

The result was one name: Morten Hjulmand, born 25 June 2026, then twenty-one, playing for a small Austrian club, Admira Wacker. He appeared in no young-talent ranking. He appeared in no transfer bulletin. He appeared only in a database I built myself, because he satisfied exactly one condition I considered important: few minutes, heavy pressure.

I wrote a forty-seven-page report on his strengths, weaknesses and integration potential, and sent it to three major clubs. One replied. Two years later, Hjulmand moved to Serie A, and later to a leading Portuguese club. My report was cited as an example of foresight. But I know better than anyone: what I had was not foresight, it was a question narrower than the industry is used to asking.

Missing data is not useless; it is a map pointing to where nobody has measured. When a metric does not exist, that is usually a sign nobody thought it worth measuring. Data gaps in sport are distributed very unevenly, and that asymmetry is the advantage. Where there is no data, there is the least competition, because most people in the trade simply follow paths that already have metrics attached.

We do not need more data. We need better questions so that old data can speak. The same dataset on minutes played and pressures applied can yield opposite conclusions, depending on whether we ask it about talent or about luck. Over years in this work, I have learned that the cost of one more data source is always higher than the cost of thinking of a better question, yet the industry always chooses to pay for the data source.

The price of waiting for perfect data

In the 2026-23 season, I led transfer strategy for a second-tier club in Boston. My number one target was a Brazilian full-back. I pursued him across three transfer windows, building an analytical framework covering technical metrics, physical metrics, injury history, even family circumstances and cultural fit. My budget was 2.4 million dollars.

While I was completing the model, another club closed the deal within forty-eight hours. They had less data than I did, fewer scenarios, fewer questions. They had one advantage: they decided first. The board told me plainly that a perfect model does not exist, and that in the transfer market timing is a variable weighted as heavily as player quality.

That lesson was expensive. The true value of a deal only becomes visible when the market goes quiet. But to have a chance of seeing that value, you must still be in the game when the noise ends. An analysis that is right but late is worth zero, or less, because it consumes the time of people who should have been acting.

Since then I have rebuilt my process around three rules. First, every analytical scenario has a hard deadline, set by the market, not by me. Second, I define the exit condition in advance: if after a set number of review rounds the data still cannot separate two options, I choose the one that is cheapest to correct. Third, I separate two kinds of risk: the risk of a wrong decision that can be fixed, and the risk of deciding nothing at all. The second is rarely entered into a spreadsheet, but it is the most expensive.

In sports commentary, the topic of delayed decisions is almost never raised. People argue about whether a player is worth the money, rarely about how much a club lost by deliberating too long. But in internal files, that is usually the largest and hardest line item to justify.

Wasted runs and the art of packaging metrics

Based on my experience watching matches across many competitions, one family of metrics is always presented as a measure of effort: distance covered and sprint counts. They appear steadily in bulletins, in player comparisons, in post-match analysis. They have a great advantage for writers: they are always positive, always legible, and always allow the conclusion that one player works harder than another.

The problem is that high distance covered can signal two opposite things. It can signal a midfielder with good coverage, reading the game early and moving ahead of the ball. It can also signal a player positioned badly, arriving late, then running to catch up. The same 12.3 kilometres can belong to the player controlling the middle, or to the player constantly chasing the ball.

When I worked with positional data at club level, I usually paired distance covered with two other metrics: average distance to the nearest teammate when the team has the ball, and the number of times the player was first to reach a dropping-ball zone. Only when those two move in the same direction as distance do I treat it as effective effort. Otherwise it is wasted running, and wasted running also produces very pretty numbers.

This packaging mechanism is not a failure of data. It is a failure of how metrics are chosen to tell a story. Sport has seen the same thing in esports, where damage or kill-participation figures are used to judge a support player, when that role largely consists of creating space, denying vision and holding tempo. Those contributions do not show up on the scoreboard, so they are easily treated as non-existent.

This leads to a very concrete market consequence. Players with pretty effort metrics are valued above players with modest ones, even when the real contribution to match outcomes is the reverse. And because valuation drives behaviour, youth academies start training players to run more rather than to run correctly. The metric does not reflect reality; it shapes it.

In scouting work, I always ask the reverse question of any effort metric: if this player ran twenty percent less, would the result change? If the answer is no, the metric is measuring presence, not value. Most bad transfer decisions come from confusing those two things.

Crisis as a demolition contractor

In 2026, when competitions paused because of the pandemic, I was a mid-level employee responsible for the financial model at a club in Massachusetts. The season was cancelled, revenue stopped, and the board needed a plan within two weeks. I built three scenarios for restructuring the contracts of key players, based on ten seasons of fan-retention data.

The result: the club saved 1.2 million dollars in wages over six months. But one of its leading players was sold because of internal conflict during negotiations. It took me another four months to convince the board that the long-term consequences of selling that player outweighed the immediate savings. I lost that argument, and I still believe I was right.

Crisis is not the enemy of the industry; it is the demolition contractor for what has already rotted. The pandemic did not create the weakness of clubs living on sponsorship bubbles; it merely made that weakness visible sooner. Much of what was torn down in 2026 was structurally unable to stand on its own for years before, and nobody wanted to say so while money was still flowing.

From that experience, I write about crisis in the language of scenarios: causes, variables, recovery paths. I habitually simulate multiple scenarios in every analysis, from the most optimistic to the most pessimistic, to test how durable an argument is. A conclusion that only holds in the optimistic scenario is not a conclusion; it is an assumption in make-up.

What I learned about data during that period is also worth remembering. We had very little new data, because the season had stopped. The entire model was built from ten seasons of old data. It worked, not because old data is better, but because we were forced to ask more precise questions of what we had. When you cannot buy more data, you are forced to think.

Ex-star academies and the gap at grassroots level

Over years of watching youth development projects in both football and esports, I keep seeing a striking pattern. A retired star opens an academy bearing his name, holds a launch event with media coverage, signs a few memoranda with local schools, and announces a mission to develop the next generation. Three years later, most of those academies operate as fee-paying service centres for well-off families, not as talent-detection facilities.

I do not deny the value of a former player passing on experience. But elite playing experience and youth coaching ability are two different skills, and only one of them was validated by the founder's career. The real gap lies a level lower: grassroots coaches. In many countries, the person teaching football to ten-year-olds is an uncertified volunteer, while the person who once scored in a top-flight league is paid to stand in front of a camera.

The system does not create genius; it only creates space for genius not to be suffocated. If I had to allocate a youth development budget, I would put most of it into training and paying grassroots coaches, and a small part into facilities. Pitches can be bought. Someone who can teach a nine-year-old to observe before receiving the ball is far harder to buy.

In esports, this problem has its own variant. Youth teams are often built around a famous retired pro, and selection criteria are frequently dominated by individual ladder rank rather than learning capacity. The result is a very narrow talent pipeline, drawn mostly from players who had the means to grind many hours a day early on, not from the players with the best game-reading ability.

This brings us back to the starting point. When a talent-detection system measures only what is easy to measure, it will find only what is easy to see. Players with real value often sit in the zone where metrics have not yet reached, and finding them requires a new question, not a new dataset.

The counter-view: honesty has a price too

I have spent most of this piece defending one principle: do not conclude without evidence. But I have to argue against myself, because that principle has a blind spot.

For big clubs, refusing to conclude is a rational choice. They have budget, time and other opportunities. For a second-tier club with 2.4 million dollars and one season to save its position, saying the data is insufficient usually means doing nothing, and doing nothing is a decision with consequences. Honesty about data becomes a privilege of the rich.

What we call a "star" is often just the person who appeared exactly when the system needed him. But in the other direction, what we call "thorough analysis" is sometimes just the analyst who appeared exactly when the organisation could afford to wait. Both are products of circumstance, not of personal quality.

There is another issue the industry rarely discusses. The market does not pay for accuracy; it pays for decisiveness. A report full of blank cells will not be shared, will not be cited, will not be commissioned again. A wrong but confident report will be cited thousands of times. This incentive structure explains why that empty report is the exception, not the norm.

The way I balance the two is by separating two kinds of questions. For questions that can wait, I keep the rule of not concluding. For questions that cannot wait, I conclude early but state the confidence level, the variables that could make the conclusion wrong, and the conditions for revision. A conclusion annotated with its error margin is more useful than a perfect conclusion that arrives late. More importantly, I do not pretend to be in the position of a big club while working for a small one.

What remains when the noise settles

The next competitive advantage in sport will not come from collecting more data. Statistics platforms have already done that, and most of the important data already exists somewhere, publicly or semi-publicly. The advantage will come from two harder things: the ability to ask the right question of data you already have, and the ability to decide fast enough that the decision still has value.

For fans, this has a direct consequence. Most of what you read in the transfer window, most of what you hear in analytical bulletins, is not built from evidence. It is built from the need to hear an answer. Knowing how to tell those two apart is the most useful skill a sports follower can develop in the coming decade.

If the next report you read had to leave three cells blank because there was no data, would you dare leave them blank? And if you were the one commissioning it, would you pay for that blankness?