The Empty Spreadsheet: Nine Dimensions of Table Tennis Analysis and the Discipline of Saying 'Insufficient Data'

02:14 in the morning, the fourth day of the third week of the transfer window. The third monitor in the corner of my Shanghai office produced a file of zero kilobytes. The filename: stage-one deconstruction output. I opened it. Blank. Not a single line.

Above it, the nine tabs of the analysis template were still open, arranged in two rows. The first tab was waiting for technical, tactical and equipment data. The second was waiting for player data and head-to-head records. The third was waiting for the event system and ranking rules. And so on to the ninth, waiting on the transmission chain of an entire industry. All of them empty.

Across twenty-nine years of watching this sport and nineteen years of working with data, I have learned one reflex that runs against instinct. When the sheet is blank, the first job is not to fill it. The first job is to classify the blankness: blank because nobody has measured it, or blank because there is nothing to measure.

That night, the answer was the second kind. And the correct article for that night was an article that does not exist.

At the desk I run a two-stage pipeline. Stage one breaks raw text into discrete information points: subject, action, timestamp, figure, source. Stage two takes those points, checks them against a fixed nine-dimension template, and is only permitted to conclude if it can name the exact information point that generated the conclusion. The template was not built in an afternoon.

In 2026, aged thirty-six, I published an analysis of a well-known foreign striker at a Shanghai club. He scored eighteen goals that season. When he started, the team's PPDA was 14.3 — loose beyond belief. When he sat on the bench, it was 9.8. I wrote that he was a defensive obstruction at the front of the pitch, a man hiding behind his goal tally. The internet called me a bookworm. A month later that club lost 0-4 to a direct rival, and the first goal came from a failed press by that very player. The old piece was dug up again.

In 2026, before Germany met South Korea in the group stage, I built a model on two variables: the retreat speed of the back line and the number of sprints above 25 km/h. The model gave Germany an xG of 1.8, but raised their probability of losing to 22 percent because the centre-backs pushed too high. I wrote two thousand words under the headline The German Machine Is Rusting. The experts laughed. South Korea won 2-0. The piece was shared more than fifty thousand times. The Korean shock was never a shock — it was the first time the number was listened to.

In 2026, with stadiums empty, I collected data from 312 Bundesliga and Premier League matches. Home win rates fell from 46 to 38 percent. Yellow cards for away teams dropped 27 percent. I wrote ten thousand words titled The Crowd Is a Statistical Variable. The pandemic did not create an exception; it exposed a rule that had been waiting all along.

Those three episodes taught me one thing: a conclusion must never walk ahead of the data. In statistics, the failure to find a significant difference is called a null result. A null result has a name, has value, and most importantly — it can be published.

My nine-dimension template divides every match, every player and every tournament into nine groups of questions. Each group has mandatory inputs. Without inputs, the group is marked blank, and every sentence that depends on it is struck from the draft.

Dimension one — technique, tactics, equipment. This group asks what kind of table tennis a player is currently playing, and whether it works. Four indicators: technical progress against their own form three months ago; execution efficiency, measured by point-win rate and rally-win rate in decisive situations; physical fit, meaning whether the movement still carries enough safety margin once hips and shoulders tire in the fifth set; and the lock data — point-win rate on own serve, rally-win rate on receive.

Table tennis has its own sub-branch here: equipment. A player who changes rubber or blade enters an adaptation window of several weeks, and in that window spin on serve and placement control are the first things to drop. A writer without spin measurements can only describe feelings. And feelings are what the naked eye sees, not what the ball carries.

Dimension two — player data and head-to-head. World ranking, points-defence pressure, and the fit between ranking and true strength are the first three columns. The fourth is head-to-head, split into three tiers: all-time, last two years, and the three majors alone. A handsome all-time record can hide an ugly two-year record, and vice versa. The fifth column is the nemesis flag: is there an opponent whose style locks this style down, regardless of ranking. The sixth is win rate abroad. The seventh is major-tournament consistency. The eighth is performance in deciding sets and in high-weight points.

Truls Moregard reached a world championship final in 2026 without being seeded. To the naked eye, that is a fairy tale. To this set of indicators, it is a case that needs rereading: a young player with an unusually high rally-win rate in the third beat of a rally — a rate invisible on television, because television only broadcasts the long rallies. A player's value does not lie in the celebration, but in the square metres of table he covers in the first three beats.

Dimension three — event system and ranking rules. Points for the champion, prize money, field strength, and where the event sits in the Olympic cycle. In table tennis this dimension also covers the draw: which half is harder, whether a nemesis could arrive early, and whether same-association separation was actually enforced. A tournament misread here produces false conclusions about form.

Dimension four — competitive landscape. Four tiers: the leading tier, the chasing tier, the emerging forces, the rest. Three comparison indices: seats in the world top ten, titles at the last five editions of the three majors, and depth in the under-21 cohort. China has held the advantage on the first two for years. The third is the one worth watching, because it measures the future rather than the past. Japan with Tomokazu Harimoto, Germany, Sweden and South Korea form the familiar chasing group. Brazil with Hugo Calderano, and a handful of European associations, are the emerging forces capable of reshaping the draw.

Dimension five — rules and governance. Every playing-rule reform, every change in the event system, every adjustment to selection criteria, every disciplinary sanction must be read across four columns: who benefits, who loses, which historical precedent is comparable, and how long the lag is before impact. Selection disputes generate the most ink, because they sit at the intersection of quantitative criteria and human judgement. One set of criteria, two readings, two outcomes.

Dimension six — coaching staff and talent pipeline. The head coach's competence and authority. The fit between player and personal coach. The stability of the staff. The age structure of the main squad. Conversion efficiency from the youth cohort to the senior team. The core structure inside the team. Signals about which young players are being trusted. Pairing strategy. In a team with a long tradition, this is the most neglected dimension, because it produces no highlights.

Dimension seven — risk surface. Six groups: competitive risk, selection risk, generational-gap risk, governance and public-opinion risk, systemic risk, and risk from opponents. Each carries a probability, an impact level and a mitigation. A risk table only has value when it dares to leave cells empty.

Dimension eight — public narrative and expectations. A narrative only endures when fundamentals back it, and those fundamentals must survive a sample-size check. This dimension measures the gap between market expectation and objective assessment across three fronts: player results, matchup outcomes, and selection outcomes. Social-media heat divided by fundamentals is a frightening ratio, because it exposes figures pushed upward by emotion rather than by results.

Dimension nine — industry transmission. Upstream is equipment, youth development and training. Midstream is events, associations and clubs. Downstream is broadcasting, commerce and derivative markets. A change upstream — the price of rubber sheeting, for instance — takes a few quarters to reach downstream, but once it arrives it moves the commercial value of players. This chain is why I watch rubber prices and world rankings with the same attention.

Those nine dimensions are the template. It is not clever. It is only stubborn.

That stubbornness shows in one small detail: in the template, every conclusion must declare its origin. If it cannot, the sentence is deleted. In a four-thousand-word piece, this operation can remove a third of the length. That is why many of my articles end up shorter than planned — and why those articles are never dug up later for the wrong reasons.

That night, what I had to resist was not the absence of data. It was the professional reflex: filling a blank cell with a sentence that sounds reasonable.

In this industry, analysis is not scarce. Silence is. A piece can look thoroughly professional, stuffed with adjectives about character, hunger and moments of brilliance, and carry not a single supporting number. That kind of writing has a dangerous property: it is not wrong, it is merely unverifiable. And what cannot be verified cannot be refuted — which means it lives on, accumulates, and eventually hardens into collective prejudice.

I write dryly, so that the game we love is not buried by the emotional hand.

There is a subtler temptation: aestheticising the void. With no data, a weak writer turns the shortfall itself into an essay about the mystery of sport. It sounds literary. But it is a lie with a beautiful form: it converts what cannot be measured into what cannot be measured at all, and converts the writer's limitation into the nature of the game. Those are two entirely different things.

Distinguishing correlation from causation belongs here too. Once a match ends, the brain automatically stitches two discrete events into a causal chain, because people need a story more than they need a spreadsheet. A player changes rubber and then loses three matches: correlation. To call it causation, you must first eliminate the time variable, the opponent variable and the fitness variable. Labelling those two words correctly in every article is a discipline, not a rhetorical trick.

The same line of numbers runs across two arenas: football and esports both bow to the algorithm. But the algorithm only bows to clean data.

Looking back at those nine dimensions open on the screen that night, one thing struck me: the value of the template is not that it helps you conclude faster. It is that it gives you a legitimate way not to conclude. In an industry where everyone must have an opinion before the first ball is served, the ability to say "not enough" is an asset, not a gap.

Ma Long and Fan Zhendong have been placed side by side in countless comparison pieces for a decade. Most of those pieces measured nothing; they merely rearranged reputations. A nine-dimension template answers that question with top-ten seats, with third-beat rally-win rates, with the under-21 depth of the two table tennis systems behind them. Reputation can be argued with. Indicators cannot.

That night I saved the blank file, named it by date, and published nothing. Three weeks later the data source was restored and the template ran again from the start. The result differed sharply from anything I could have written that night — but that belongs to three weeks later, not to that night.

Transfer season is the season of informational gluttony, and the season in which a numbers writer is most likely to err. The only defence is a protocol capable of returning a null value. A system that can only conclude is an incomplete system; a system that knows how to stay silent has begun to be trustworthy.

The naked eye will keep seeing what it wants to see. The job of the numbers writer is not to argue with it. The job is to record the timestamp, record the indicator, and let the trace speak when its moment arrives.

When the naked eye sleeps, the data stays awake — and it saw it coming. But when the data sleeps too, the most honest thing to do is turn off the light and wait.