Eleven Tables, Not One Line of Data: The Hardest Discipline in Football Analysis

Eleven Tables, Not One Line of Data

Last July, in Chengdu, I received a nine-page dossier about a club playing in China's second tier. It came from four freshly graduated analysts who had worked carefully and laid it out beautifully. The dossier had eleven tables. A six-row risk matrix. An industry transmission diagram running from youth academies to broadcast rights. A squad-value comparison against four direct rivals. A public-pressure table broken down by head coach, senior players and board.

I read every cell. Every cell had text. But in the conclusion column, every row said: insufficient information to assess. Eleven tables, not one line of real data. The template was so flawless that if you only skimmed it, you would have signed it off without hesitation.

An empty table, properly formatted, is more dangerous than a wrong table, because it never triggers the reader's suspicion. A wrong table gives you something to argue with. An empty table that still looks good only earns a nod.

We start from Chengdu, where the barriers are not as towering as people assume.

In 2026, at forty-one, I sat down and reconstructed fourteen passing sequences from a second-tier match to prove the away side had a fixed blind spot behind its two full-backs. I was told to my face that women do not understand high pressing. That video reached 120,000 views, six times the official channel. From then on I adopted one rule: every tactical claim needs at least three concrete in-match situations behind it.

When I started auditing other people's dossiers, I realised that rule had a hole. It guards against the wrong. It does not guard against the empty.

Why a table became free

Fifteen years ago, producing a data report on a second-tier match required someone to sit and chart it, software, and time. That cost naturally filtered out meaningless reports. Today it is entirely different. A second-year student can build a full dashboard in an afternoon — metrics, trend lines, conclusion column, even a source note — without watching a single match.

That is why I noticed in July. Value has shifted. The scarce thing is no longer the ability to build a table. The scarce thing is the discipline to leave a cell empty.

Put another way, my job has changed from measuring a lot to refusing to measure without a basis.

So that readers unfamiliar with this vocabulary can follow, let me state a few fundamentals plainly. PPDA is the number of passes a team allows its opponent per defensive action of its own. The lower the figure, the higher the pressing. xG is expected goals, a model estimating chance quality from shot location and situation, entirely separate from actual goals. Two lines of definition are enough to expose the problem: both are estimates, and both are very easy to narrate wrongly.

I know plenty of people who look competent still need to hear those definitions again. That is fine. Definitions are the one thing that cannot be inflated.

Take a traceable example. On 6 February 2026, the Premier League announced charges against Manchester City for 115 breaches of its financial rules, per documents released by the league. On 17 November 2026, Everton were docked 10 points for breaching profit and sustainability rules; the sanction was reduced to 6 points on appeal on 26 February 2026. On 18 March 2026, Nottingham Forest were docked 4 points. Those three dates appear in thousands of analytical tables worldwide, and most of those tables add nothing, because whoever built them never opened the original filings.

That is the mechanism. Data leaves its source, enters a table, and becomes decoration.

What we call a "system" is really the product of a thousand repeated training sessions

Russia 2026 gave me a concrete lesson I still use as a template. In the quarter-final between France and Uruguay on 6 July 2026, Didier Deschamps dropped Antoine Griezmann deeper, turning a nominal 4-3-3 into a 4-4-2 variant out of possession. The distance between France's midfield and defensive lines held around 19.5 metres throughout the first half. Edinson Cavani almost never received the ball with his back to goal in the space between those lines. I filed the analysis that night, ahead of the European press, and it was shared 40,000 times.

What I did not write that night, and now regret, was a caveat about the limits of that 19.5-metre figure. It was true for that match. It is not true for another. A metric says nothing on its own; it only says something when placed beside another metric from the same moment, in the same ball state. Low PPDA without a block-shape indicator is a bare number. High xG without shots inside the box is an illusion.

When I reviewed a second-tier match last March, the home side published a 4-2-3-1 but operated as a 4-4-2 out of possession. In active defensive phases, the gap between their lines hovered around 19 metres. In attacking transition it compressed to roughly 12 metres. That seven-metre swing decided the whole match, because it was exactly the space the opposing midfielder used to receive on the half-turn. No table tells that story unless the analyst sits through fourteen sequences.

Tactics are what you use when the opponent thinks they have read you. Data is what you use when you yourself need to know where you are reading it wrong.

The counter-intuitive angle: an empty cell is a conclusion, and sometimes an excuse

Russia 2026 taught me that attack is a form of expression and defence is the answer. 2026 taught me that a team stands on a system, not a line-up. Empty stadiums are the largest laboratory modern football has ever had. Those three years hardened me into someone who leans defensive in almost every argument.

Which is exactly why I have to warn myself first. A defensive bias is an empty table in disguise. When I write "insufficient data to conclude", sometimes I mean it, and sometimes I am dodging work. Readers are entitled to tell those apart.

The tell is this: if the data really is missing, I must state three things — which data is missing, how much is missing, and which way the conclusion would move if it existed. Fail to state those three and the empty cell is just a polished excuse.

The biggest risk in my job right now is letting an empty conclusion enter a report and be read as "no issues found". Those two statements are entirely different. One says: examined and found safe. The other says: never examined. In football, the distance between them is often exactly the number of points that relegates a club.

The paradox of the empty cell — In football, silence in the data is usually read as calm in the data. A defender is never named in the highlights because he made no error. That does not mean he played well. He played so well that the opponent had to prove too many other things. An empty cell works the same way: it may be safe, or nobody may ever have opened it.

Checking back next match

Based on my experience following matches, I have a new habit whenever a report lands. I ask one question only: which cell in this table, if deleted, would change my decision? If none would, the table needs no further reading. If one would, I ask next: what data fills that cell, from where, dated when.

Cultural barriers are not dismantled by words. They are dismantled by the first match. The same holds for data. There has to be a first match, fourteen concrete sequences, a seven-metre gap between the lines. Only then does a table earn the right to exist.

The thing I want to bring into this week's review session: in your table, how many cells are coloured in because you have data, and how many are coloured in because you did not want to leave them white?