Nine Empty Fields: The Limits of Sports Analysis

The wall clock in the small Shenzhen apartment read 23:40. June rain hammered the glass, water running down the frame like a crowd clapping off-beat. A mechanical keyboard clacked in a room where only one person was still awake. On the screen sat a spreadsheet with nine tabs. All nine tabs were empty.

The first tab was called "Patch & Meta". No version number. The second, "Tournament Format". No event name. The third, "Roster & Players". No names of people. So it went for the remaining six: club finance, rules and governance, risk profile, public narrative, industry transmission. Every cell carried the same grey line, repeating like a refrain: insufficient information.

The deadline was six in the morning. I had six hours to write a sports analysis, and not a single fragment of data to start from.

Anyone who works in this trade knows this is the most dangerous moment. Not because there is too little writing to do, but because a lack of data usually produces more writing than an abundance of it.

I remember November 2026, when Saudi Arabia beat Argentina 2-1 in Lusail. That night I sat in front of a different set of numbers, and my model returned an xG of 0.35 for the winners and 1.9 for the losers. The figure said something very specific: the winning side scored twice from almost no meaningful chances. It said nothing about where, at which minute, or why Argentina's defence lost its shape. Readers reacted sharply. Some felt the number insulted the victory.

What I learned that night was not about xG. It was this: an analytical table is worth exactly as much as the data poured into it.

Tonight, my table was empty.

A professional sports analysis workflow, whether in football or esports, runs through two layers. The first is extraction: gathering events, numbers, names, tournaments, dates and sources. The second is interpretation: placing those fragments side by side, hunting for causality, building scenarios, quantifying risk. Without the first layer, the second is just prose.

The framework in front of me was built around nine dimensions. First, patch and meta — the game version and the direction in which the dominant playstyle is shifting. Second, tournament system and format — number of matches, group structure, qualification path, schedule density. Third, teams and players — paper strength, role fit, chemistry, bench depth. Fourth, the regional map — the balance of power between regions and the flow of talent. Fifth, finance and business — revenue structure, salary expenditure, transfer deals. Sixth, rules and governance — competitive integrity, registration rules, disciplinary precedent. Seventh, the risk profile — probability, impact, mitigation. Eighth, public narrative — the media heat cycle and the gap between expectation and reality. Ninth, industry transmission — the knock-on effects on publishers, broadcast platforms, sponsors and derivative markets.

Those nine dimensions make a good frame. They force the writer to answer hard questions instead of gliding past them. But a good frame is also a good trap, quite literally: an empty frame always creates pressure to be filled with whatever happens to fit.

I have followed matches in both of these worlds long enough to know that a weak writer is not a writer who is wrong. A weak writer is one who cannot tell the difference between "I know" and "I suspect", and then presents both in the same tone of voice.

So in the six hours remaining, I will walk through those nine dimensions, and at each one I will point out exactly what is missing, and what happens to the trade when it is absent.

Patch and meta is the easiest dimension of all to fabricate.

The reason is simple. Patch notes are the closest thing esports has to a controlled experiment. A publisher publishes numbers and an effective date, and within twenty-four hours millions of matches are logged on the servers. A champion's win rate can jump from 47% to 53% on the strength of a single line in an update. That is data you can verify, repeat and rebut.

When there is no version number, no effective date and no declared tournament server, every statement about a "meta shift" becomes astrology. People say Team A got stronger because their style suits the new meta. But stronger where, at what pick rate, in which event, on which patch — nobody answers, and that is precisely why the sentence exists.

Football has its own patches, they just arrive more slowly and more quietly. The five-substitution rule entered widespread use from 2026, and it changed how teams plan for the second half. The stoppage-time directive at the 2026 World Cup pushed match duration to levels never seen before; some games had more than twenty minutes of added time, which directly affected how teams managed late-game fitness. These are changes with documents, dates and citable text.

The difference between an analyst and a commentator lives right there. An analyst cites the document. A commentator cites the feeling.

In my spreadsheet, the "Patch & Meta" tab has no version number. Which means I cannot say anything about the direction of travel. And if I say it anyway, I am selling readers a belief dressed up as a fact.

Tournament format is the dimension where every conclusion depends on sample size.

A win in a three-match group stage does not carry the same meaning as a win in a best-of-five. This is basic arithmetic that very few analyses bother to do. With three matches, the confidence interval around any metric is so wide it becomes useless. With seven matches in a Swiss format, you begin to see a shape, but still not enough to assert. With a best-of-five, you have a small sample but a structured one — and structure matters more than size.

The Swiss format was introduced into a major title's world championship in 2026. It completely changed how results should be read. Under the old format, a team could go far on an easy bracket. Under Swiss, a team's fate depends on whom they meet at which record, and that creates a systematic unfairness no standings table displays.

Euro 2026's expanded format created a similar distortion on the football side. More teams advance, which means a win on the final matchday can carry a side through despite two dreadful opening games. Anyone who reads the table without reading the format will draw the wrong conclusion about that team's true strength.

For VCS — Vietnam's top-tier league in that discipline — the season structure, the number of teams and the allocation of international slots are the variables that determine the value of any conclusion about a player. A player with a beautiful group-stage stat line is not stronger than a player with a modest knockout-stage one. Two different samples, two different stories.

In my spreadsheet, the "Tournament Format" tab has no event name. Which means every cross-event comparison is meaningless. I could write three thousand words about it, and all of it would be prose.

Roster and form is where data gets abused the most.

The form curve is a beautiful concept and largely an illusion. Over five matches, a player's attacking output fluctuates mostly because of noise. You need fifteen to twenty matches before you can start separating signal from static, and in esports very few tournaments give you that many matches on a single patch.

That is why I always place a note about sample size next to every metric. Without that note, a number becomes a declaration.

Georgia at Euro 2026 is the example I keep returning to when I teach interns. In qualifying, that team conceded an average of roughly 0.9 expected goals per match, among the lowest in the field, even though they did not dominate possession. That was a long enough sample, on a stable tactical system, with the same group of players. When they beat Portugal 2-0 in the group stage, the result was not outside the model. It was inside the model. The model simply had not been read.

On the Vietnamese side, Nguyễn Quang Hải, Nguyễn Tiến Linh and Đỗ Hùng Dũng are all cases where goals and assists do not tell the whole story of a role. A playmaking midfielder can go a whole tournament without scoring and still be the man who sets the tempo. A striker can score four, three of them from the penalty spot. If you do not separate types of goals, you have merged two different jobs into one number.

I spent a month re-watching footage from a 2026 World Cup semi-final to understand why my model understated reality. France beat Belgium 1-0 through a Samuel Umtiti header from a set piece. France's raw xG that night sat around 1.6, Belgium's around 0.8. The model was not wrong about open play. It simply had no weighting for a phase of play in which everything is decided inside one second by a player whose only job is to arrive in the right place.

When Lionel Messi or Kylian Mbappé produce a moment, the model logs it as a shot. But that moment was created ten seconds earlier, by the positioning of seven other players. Tracking data captures that. A basic stats table does not.

In my spreadsheet, the "Roster & Players" tab has no names. No names, no roles, no form curves. Just a rectangular blank.

The regional map is the dimension where data has to cross borders.

The balance between regions in esports is measured by three things: international head-to-head results, the depth of the talent pool, and the output of youth development systems. All three require match-level and player-level data stretching over years.

The major regions hold a structural advantage in that they play each other year-round. When a region has only a handful of international slots a year, every comparison rests on a thin sample — and a thin sample always inflates the good results as much as the bad ones.

The academies of the big organisations are an example of a talent-hoarding structure. Most academy players never get promoted to the main roster. The real figure is far lower than the impression the media creates, because the media only reports on the ones who make it. This is a survivorship bias programmed into the way the industry tells stories about itself.

In 2026, VCS went through an integrity review that led to multiple players being suspended. As a data journalist, the striking part was not the list of sanctions. It was this: the league's metrics from that period need to be re-read, because some matches may no longer reflect the teams' true ability. Old data does not correct itself when new facts emerge. The reader has to do the correcting.

In my spreadsheet, the "Regional Map" tab contains no regions. Which means every comparison of LCK with LPL, or VCS with LEC, is a word game.

Finance is the dimension where every number is a life converted into currency.

An esports team's revenue structure rests on sponsorship, publisher revenue sharing, jersey sales, and in many cases direct owner funding. Salary expenditure is the biggest unknown, because esports contracts are largely private. Without numbers, every analysis of roster strength rests, indirectly, on rumour.

Football offers a clear comparison because contracts there must be disclosed in many league systems. Neymar's 2026 transfer from Barcelona to Paris Saint-Germain, at a fee of 222 million euros, reset the entire pricing floor of the market for years afterwards. But the interesting part is not the figure. It is that the figure forced dozens of other clubs to reprice their own players, setting off an inflationary cycle no one controlled.

Every transfer number is a life converted into currency. Behind the number is a twenty-year-old who has to move house, move language, move the very way he plays. No balance sheet records that cost.

In my spreadsheet, the "Finance & Business" tab is empty. Which means I cannot say which team is richer, and I cannot say which team is close to collapse.

Rules and governance is the most dangerous dimension to leave blank.

Here, empty space is not neutral. It protects the things that need scrutiny. The checklist includes competitive integrity, transfer and registration rules, contract compliance, protection of minors, and disputes between teams and publishers.

When there is no data in this dimension, the correct response is not silence. The correct response is to state plainly that the data is absent, and to describe the worst-case scenario if a given assumption turns out to be true.

The worst-case scenario is not a sanction. The worst-case scenario is prolonged opacity, where the public cannot tell which matches are trustworthy and which are not. That is a loss no verdict can compensate for, because it destroys the thing every league lives on: the belief that results are real.

A risk profile without data is just an empty matrix.

The risk matrix has six categories: competitive, financial, personnel, rules, public opinion and systemic. Each needs three things — probability, impact, mitigation. Without data, none of the three exists, and the only honest move is to write "cannot be assessed" in every cell.

This is where many analyses choose a different path. They fill the cells with "low", "medium", "high", then close with an overall rating. The analysis looks complete. It also looks like a usable product.

But a risk rating with no data behind it is no different from a weather forecast written before looking out of the window.

Public narrative is the dimension I watch most closely.

Every team has a story told about it, and the gap between that story and measurable reality is the expectation gap. It is a metric that appears in no statistics table, yet it governs how fans read every other table.

Lee Sang-hyeok, known as Faker, is a case study in the media heat cycle. Every time he reaches a final, search volume spikes, and every time his team wins, the story told stops being about the match and becomes about endurance itself. That creates a side effect: the other players on the same roster get pushed into the background, and their numbers get read through the shadow of one name.

The media heat cycle has a very regular shape: a big event, a peak week, a declining week, then a new story replaces it. The professional question is whether the story has a data foundation, and whether the sample is large enough for it to survive into the next cycle. Mostly it is not. Most sports stories live about ten days.

In my spreadsheet, the "Public Narrative" tab is empty. No heat cycle, no sentiment reading, no expectation gap. Just a cell where I know for certain that I have measured nothing.

Industry transmission is the last dimension, and the one most easily abused to end an article.

The transmission model runs from publisher, through broadcast platforms, through sponsors, through derivative markets, to the mainstreaming of the discipline. Each link needs its own data: concurrent viewership, rights fees, sponsorship budgets, the number of training facilities, the legal framework in each country.

When none of those links exist, the closing section on industry transmission becomes a speech about the future. It is always right because it says nothing specific. And it is always easy to write, because it needs no number at all.

In my spreadsheet, the ninth tab is empty, and that is the tab I miss least.

Six in the morning. I look again at the nine tabs. All still empty.

This is the point where I have to decide something about my own trade.

There is a version of this job in which I fill the nine empty cells with nine plausible paragraphs. I open with an observation about a meta trend, cite a few prominent teams as examples, add a finance section with estimated figures, and close with a soft prediction. The piece reads smoothly. It gets shared. And it is not wrong in any way that can be caught, because it asserts nothing specific enough to be caught.

That is the worst kind of content in this industry, and it is also the most produced.

What I want to state clearly here is a counter-intuitive position, and it is not an easy one to hear for people who do this work — including me.

The temptation of a complete framework is a greater enemy than ignorance.

The nine-dimension frame was designed with no gaps. When you see a table with nine cells, your instinct is to make it full. That is a craftsman's instinct, not an analyst's. A craftsman completes the product. An analyst completes the verification.

Modern sports analysis has built an ecosystem that rewards volume of content more than accuracy. Every day, thousands of articles are published about matches that have not yet happened, based on models the writers cannot verify. The number comes first, the source is checked afterwards, and if no source can be found, the phrase "according to some statistics" settles the matter.

This is where I have to say the thing I have said to myself for years, and sometimes still need to repeat.

xG does not lie; it simply never tells the whole truth.

The same principle applies to every metric in those nine tabs. Patch notes do not lie. Standings tables do not lie. Contracts do not lie. But all of them leave a gap in the middle — a gap the storyteller's skill must step into, and a gap the storyteller's laziness must be blocked from.

I once fell into the inverse trap: paralysis through over-verification. There were weeks when I added a third source, then a fourth, for a number the first two had already agreed on. I told myself it was caution. In reality it was procrastination disguised as professional ethics.

The fix I have applied since is a quota: two independent sources per key figure, then write. If I find an error after publication, correct it and date the correction. An article corrected in public is worth more than an article that never existed.

And which data cannot measure this moment?

That is the question I force myself to answer at the end of every piece. For tonight, it has a very clear answer: my nine tabs can measure everything except one thing — whether readers trust me. Trust is not in the spreadsheet. It is in how I handle the empty spreadsheet.

I choose this: publish the empty table as it stands, alongside a map of what would be needed to fill it.

Specifically, for these nine tabs to become a real analysis, I need a minimum list. For the patch column: version number, effective date, tournament server version, and professional-level win rates over at least two weeks after the update. For the format column: event name, tier, number of teams, group-stage structure, maximum matches per team, and schedule density in the peak month. For the roster column: player list, roles, recent match counts, and a performance metric with a stated sample size. For the regional column: international head-to-head results over the past three years and each region's slot allocation. For the finance column: an approximate revenue structure and wage-payment status. For the rules column: disciplinary precedents from the past twelve months. For the risk column: a list of possible events with probability estimates stating their method. For the media column: weekly search and concurrent-viewership figures. For the industry column: data from at least two independent sources in two different markets.

Without that list, an analysis is just a handsome frame.

There is a line I tell myself every time I sit in front of an empty table, and it has become a rule of practice.

I do not build a table for the match; I build a table for the doubt.

A table for the match exists to prove something I already believed. A table for the doubt exists to find where my belief might be wrong. The second kind always begins with empty cells, and always ends with a to-do list longer than expected.

In this case, the most honest conclusion I can draw after walking all nine dimensions is this: there is no basis for any conclusion on competitive, financial or industry grounds. That is a conclusion of value. It saves the reader the time of reading an unsupported piece, and it saves me the credibility I would lose by writing one.

Every major tournament season shares this feature: the flow of information moves faster than the capacity to verify it. Bulletins are pushed out before the match ends. Numbers are quoted with no traceable origin. The pressure to have a strong opinion outweighs the pressure to have a correct one.

In that environment, writers hold a strategic advantage few choose to use: the right to say there is not enough data. That is not weakness. It is a statement of method, and a discerning reader will recognise the difference between someone who does not know and someone who knows that they do not know yet.

Whether the stadium has a crowd or not, the match still needs someone to tell it back. But the teller must be someone who was at the stadium, not someone who read another person's minutes and believed they had been there.

It is now 6:12. The rain has eased. I save the spreadsheet with its nine empty tabs, name the file with today's date and the suffix "insufficient data". Then I send my editor one line: the analysis is waiting on the extraction layer, not on me to write more.

In ten years of watching this industry, I have seen many correct predictions forgotten and many wrong ones remembered forever. The only thing that survives both is method. A prediction can be right by luck. A method is only right because it was built.

The next turn of the season will bring a new set of patches, a new set of transfers, a new set of stories with no sources. The signal I will be tracking is not which team is getting stronger. It is the question of how many analyses in the coming peak will dare to leave a cell empty, and write in it the two most honest words in the trade: not yet known.

Data is a monastery, but I choose to leave the gate to find the match. And when the match has not yet been recorded, the first thing I do is not write. It is wait.