The Empty Column in Table Tennis Data: When Zero Does Not Mean 'Nothing'

Last Friday evening, I reopened a tracking sheet I had prepared for a professional table tennis meeting. The spreadsheet returned exactly the sort of lines anyone working with data dreads: the article title read N/A, the source read N/A, the 'information points' field was empty, the 'related entities' field contained only a meaningless instruction — 'identify from the information points above' — while above there were no points at all. Time sensitivity had not been assessed. Source quality was left blank. I stared at the screen for about three minutes, then did something I should have done from the start: I checked whether the sheet was genuinely empty, or whether I was misreading it.

This is the story of an empty column. And of something the table tennis analysis trade in Vietnam often overlooks: a null result does not mean a safe result. The gap between 'no risk detected' and 'no risk assessable' is the gap between a conclusion and a failure.

My job is a transfer market administrator specializing in table tennis. Most of my time is not spent watching balls bounce back and forth; it is spent reading player datasets: service efficiency, the share of points won in the first three exchanges, effectiveness in long rallies, the 52-week rolling ranking system used by WTT, and dry indicators such as career age, tournaments entered per year, and match density.

In Vietnam, the data infrastructure for table tennis remains thin. Names like Nguyen Anh Tu, Mai Hoang My Trang, and Nguyen Khoa Dieu Khanh are widely mentioned domestically, yet systematic data for evaluating each of them is scarce. National tournaments usually publish only the final result — who won, by what score. What lies behind the score is rarely recorded. A player can win 4-0 by scoring directly off the serve, or win 4-0 because the opponent simply missed. On the scoresheet the two matches look identical. In meaning, they are worlds apart.

That is why, whenever I prepare an evaluation of a player, I build a multi-layered sheet: a raw-data layer, a tournament-context layer, a psychological layer, a head-to-head layer. Once it is built, I discover that some matches leave my dataset entirely empty — not because there was nothing worth saying, but because the extraction stage dropped all of it.

In 2026, I bet on xG. The V-League answered with a shock. That April, in a match where InStat recorded the home side generating only 0.9 xG against 1.7 xG for the opponent, the final result was a win for the team with the lower figure. The media praised the coach; I wrote a line in my notebook: an abnormally high conversion rate does not last. The team then dropped points in a string of matches. The truth lies in the numbers, but the numbers must be read correctly.

That lesson followed me into table tennis. A player winning three straight matches through the short exchanges — what the trade calls the 'first three shots': the serve, the receive, and the third-ball attack — can look ascendant. But read point by point, and the winning points may come more from the opponent's faulty receives than from the quality of spin. Two causes, one score. The person reading the scoresheet sees progress. The person reading the data sees a lucky loan not yet due.

Then the 2026 World Cup taught me: data is never a single layer. I once predicted the champion by pooling every metric across the whole tournament, only to watch my pick get eliminated while the eventual winner was the side whose defensive numbers improved markedly from the group stage to the knockout rounds. I had pooled one fixed figure across every phase, while football — like table tennis — changes how it plays from segment to segment. Layering data is how I keep my composure amid frantic evaluation cycles.

And that was precisely what I missed in that empty sheet on Friday. A blank extraction does not mean the subject had nothing to analyze. It only means we failed to retrieve the information. Yet in many professional meetings, I have watched colleagues read a null result as a safe conclusion: 'No risks identified.' That is the most dangerous error a data person can make.

In table tennis, this confusion is especially easy, because the sport's risk catalogue is far from small. Some players are mid-way through a technical overhaul — switching to a heavier topspin style, or changing rubber — with the adaptation window still open. A new rubber can ruin the feel for spin for two weeks. A small change in sponge hardness can shift a loop by a few millimetres, and at the elite level a few millimetres is a whole match.

Some players accumulate multiple events at once — singles, doubles, mixed doubles — and at some point the body collapses. There is the pressure of defending ranking points under the rolling deduction system, forcing athletes onto court more than their bodies allow. There are governance issues: selection criteria, the handling of disciplinary matters, the allocation of entry slots. All of these signals usually live in interviews, in quoted remarks from players and coaches — exactly the section a loose extraction stage skips most often. When it skips, the sheet returns N/A, and the reader assumes there was nothing to say.

Here is a counterintuitive point I want to dwell on, because it runs against the instinct of most sports journalists. When an analysis sheet returns all zeroes, the natural reaction is: 'So there is no problem.' In data logic, that is a false inference. Two entirely different sentences must be distinguished: 'no risk was detected' and 'no risk could be assessed.' The first is a conclusion. The second is a failure. Confusing them can lead to ignoring the earliest warning signals — injury, a technical-overhaul slump, internal tension, selection controversy.

Once, preparing for a youth tournament, a colleague pinned a line on the board: 'The opponent has no significant weaknesses.' I asked where his information came from. He said: 'I watched the last three matches, saw no issue.' The last three matches. Meanwhile the opponent had just changed rubber, and all three of those matches were friendlies. Transfer market administrators do not manage money flows. They manage expectations. And to manage expectations, we need to know what we are reading, rather than read what we want to see.

Another trap is the habit of treating heat maps and visualizations as a new form of fortune-telling. A blazing red block on the court does not say that a player is playing well; it only says the player appeared often in that zone. A player's real role within a tactical system — the rhythm-setter, the ball-holder, the finisher — is not in the colours. It is in the sequence of decisions, something only event-sequence data can capture. And esports taught me that playing rhythm is also a data layer: the same volume of points, allocated at different tempos, tells entirely different stories.

So when is a null result genuinely 'nothing'? I would argue three conditions. First, the raw source text must exist, be readable, not paywalled, and not truncated. Second, the extraction stage must be independently audited to distinguish 'text with no information' from 'text with information that the filter dropped.' Third, every extracted information point must be tied to a concrete source, with a date and a named entity. Without one of those three, a null result is not data. It is a void wearing the costume of data.

And that is why I value source traceability above any beautiful figure. An analysis sheet where every conclusion can be tied to a numbered information point with a source and a date — that is a sheet that can stand before a coaching staff. A sheet full of general assertions anchored nowhere — that is a sheet that cannot be published, however long it may be.

Ah, that empty sheet on Friday. When I checked again, most of the original text was still there. It simply had not been extracted. The source was intact; only the reading stage had jammed. But what made me think hardest was my own first reaction: I nearly read an extraction failure as a safe conclusion. Had I nodded to myself that night, the null result would have become a professional decision.

There are seasons only xG can read, not the eye. After seven years, I trust the silence between two numbers. And the silence between one zero and another zero can be the distance between missing an injury and catching it in time — between missing an unfinished technical overhaul and preparing it an adaptation window long enough.

The question I leave for those doing table tennis analysis in Vietnam, and for myself in the coming season: next time, when the data sheet returns all N/A, will you write in the minutes that 'there is no problem', or that 'we have not read anything yet'? At the elite level, those two sentences can be exactly one defeat apart.