Nine Layers of Table Tennis Data and the Lesson of an Empty Analysis

Forty-seven pages with not a single fact in them

Shenzhen, a Tuesday morning. On my screen sits a forty-seven-page report on professional table tennis. It has every structural feature any analytics desk could want: nine sections, each with a heading, a table, a notes line, an assessment cell, an evidence column, a risk block, a conclusion, and a glossary at the end. A document formatted so perfectly that anyone skimming it would assume it contained something.

And in every cell, in every position where a name, a figure, a date, or a match should have been, the same line repeats: "N/A — insufficient information."

I read it twice. The first time to find data. The second time to understand how a document could be so formally correct and so empty. On the second pass I realised what made me sit there longest: this is the most dangerous kind of document in sports analytics. A document with a wrong number can be argued with. A document with no numbers at all, dressed as a document with findings, cannot be argued with by anyone — because it asserts nothing to refute.

It is not wrong. It is empty. And in my trade, empty is a more dangerous state than wrong.

I have worked in this field since 2026. Thirty-six years of watching, reading, recording, and re-pricing every number that has passed through my hands. I once thought I understood every failure mode of data: noisy data, biased samples, data distorted by crowd emotion, data dead from age. That Tuesday morning I met one more — data that does not exist but has been bound into a report anyway.

From Guangzhou 2026 to a table tennis table

In 2026 I was forty-three, working as a betting analyst in Shenzhen. AFC Champions League quarterfinal, Guangzhou Evergrande against Urawa Red Diamonds at home. I used expected goals to form a read and concluded the hosts would win. I ignored two things: the positional weight of each shot, and set pieces. Guangzhou lost 0-1 at home. I lost thirty thousand yuan.

After the match I sat down and logged all fourteen unsuccessful shots, mapped each one, and found the lesson I still treat as the founding lesson of my trade: the problem was not that I had been wrong about a number. The problem was that I had been wrong about what was absent from the table. I was missing a data field. I did not know it was missing, so I read the table as if it were full.

From that point I built my own positional database, and the first rule of that database was not to collect more — it was to record clearly what I was missing. Every analysis I have written since cites at least three layers: position, timing, situation. And every one carries a warning about model error.

In 2026, at the World Cup in Russia, from that same database, I published an analysis showing Croatia were the only side among the last four with an average PPDA of 12.1 — deliberately ceding the press to counter-attack, with an 18.2 percent conversion rate of expected goals from counter-attacks. Most picked France. I picked Croatia for the final. They got there and lost 2-4.

That piece reached two hundred thousand reads. But what I carried away from it was not the reach. It was a writing structure: give the number, explain the principle behind the number, then ask what the number does not say. Croatia 2026 was not there to make you believe in miracles; it was there to remind you that probability was never destiny.

I retell these old stories not out of nostalgia. I retell them because this year I moved most of my working time to a different sport: table tennis. And this sport is suffering from exactly the disease I thought I had cured in myself in 2026.

Nine layers and one professional habit

Modern table tennis looks simple from the outside: two people, one table, a plastic ball, first to twenty-one wins. From the inside, it is the sport with the highest decision density of any I have ever analysed. An average rally lasts under five seconds. Inside those five seconds there can be four consecutive tactical decisions. No other sport produces decisions at that rate.

And no other sport has such a gap between the volume of data generated and the volume of data actually understood. The ecosystem has three tiers. The first is the event system: the International Table Tennis Federation governs and ranks, while World Table Tennis — WTT — is the commercial arm, running a tiered series from Grand Smash and Champions down through Star Contender to Contender. The second is match data: the statistical sheets published by organisers and broadcasters. The third is what nobody publishes: placement positions, tempo, psychological situations, and the way a player handles pressure at ten-all.

The third tier is where matches are decided. And the third tier appears in almost no public data table.

I work in nine layers. Not because nine is a pleasing number, but because after many years I concluded that omitting any one of them produces the same failure: a confident conclusion built on a floor with a hole in it. Those nine are technique and equipment; player data and head-to-head records; event system and points rules; the competitive landscape between China and the rest; rules and governance; coaching staff and the talent pipeline; the risk surface; public narrative and expectation; and finally industry transmission.

Data never lies — but it never tells the whole story either.

What I want to do here is walk through those nine layers using actual table tennis events, and at each layer point out where the data falls silent. That silence is what separates an analyst from someone reading a spreadsheet.

The blade speaks before the player does

Layer one is the most overlooked, even though it sits in plain sight: technique and equipment.

At the Paris 2026 Olympic Games, Truls Moregard of Sweden reached the men's singles final and won silver. Earlier, in the round of thirty-two, he eliminated Wang Chuqin, China's top seed. It was one of the biggest shocks in Olympic table tennis history.

The media explanation was predictable: spirit, nerve, a moment of brilliance. My explanation is drier. Moregard plays with a blade that is not the traditional oval — a six-sided shape with a larger hitting area and a wider sweet spot. That blade changes the contact point on wide strokes, and the contact point determines the probability of the ball landing on marginal shots.

That is an equipment detail. It appears in no statistical table. It is not an inspirational story. It is a variable, and the variable is real.

Layer one has a second branch analysts call the pips style — short or long pimpled rubber instead of smooth inverted rubber, producing flat, erratic trajectories and broken rhythm. A high-level pips player is a variable every prediction model handles poorly, because any given player's sample of matches against pips in a single season is far too small.

In Europe, Felix Lebrun of France carries a different variable: a penhold grip, a style almost extinct in modern European table tennis. He still climbed into the world's top group and helped France win men's team bronze at Paris 2026 alongside his brother Alexis Lebrun and veteran Simon Gauzy.

And here is where layer one collides with layer two. When a player changes rubber, blade construction, or sponge hardness, people ask how long it takes to adjust. There is no general answer. Some take two weeks. Some take two tournament cycles. Throughout that adaptation period, every statistic attached to them is contaminated. If you read their numbers without knowing they just changed equipment, you are reading corrupted data.

No data table has a column for "just changed rubber." That column lives in the analyst's head, or nowhere at all.

The ranking is an organised lie

Layer two is where I spend the most ink and where most people go most wrong: player data and head-to-head records.

The ITTF ranking uses a rolling fifty-two-week window. Every week, the points from the tournament held exactly one year earlier are deducted. The direct consequence is that the ranking trails form. A player in brilliant form who has not yet banked points at a major carries a ranking below their true level. A player who won a major eleven months ago carries a ranking above it until those points drop off.

I call that pressure points-defence pressure. It is a psychological variable, not a technical one. But it changes how people enter matches: someone defending points picks safer patterns, serves less aggressively, and accepts longer games. Someone who needs points takes more risk. Two evenly matched players can adopt completely different tactics purely because their points histories differ.

And here is the most misunderstood part: the head-to-head record.

A head-to-head line of two wins and one loss sounds weighty. But three matches is three samples. With three samples the confidence interval is so wide that nothing can be concluded. Even a seven-to-three record is only ten matches, and ten elite table tennis matches happen under entirely different conditions: different flooring, different balls, different rubber, different physical condition, different points pressure.

My method splits head-to-head into three tiers. Career total, to understand the stylistic history. The last two years, to read current form. And a separate tier for the three majors — the Olympics, the World Championships, the World Cup — to read the ability to handle pressure. These three tiers frequently contradict each other, and the contradiction is where the information lives.

Tomokazu Harimoto of Japan is a clean example. Born in 2026, he emerged as a teenager and has been Japan's leading men's singles spearhead for years. His head-to-head against China's core group has passed through very poor stretches and then balanced out again. Read the aggregate and you see a player who loses more than he wins. Read it by time tier and you see a player who can beat anyone in a given match but cannot hold that level across a series. Those are two completely different conclusions about the same person.

The divergence between ranking and true strength is where I earn the most and lose the most. It earns when I spot an underpriced player. It loses when I forget that a low-ranked player may simply be playing badly.

Who is allowed to be at the table

Layer three is the event system and points rules. It is the layer most fans skip entirely, even though it determines who appears where, and in what physical condition.

The WTT tier structure scales points with tier level. Grand Smash is the highest, then Champions, then Star Contender, then Contender. Alongside it sits the traditional ITTF calendar, most importantly the World Championships and the World Cup, plus the four-yearly summit of the Olympic Games.

This system produces three effects that no data table displays.

First, density. To hold a ranking, a player must enter consecutive events, travel across continents, and compete in multiple disciplines at the same tournament — singles, doubles, mixed doubles. Each added discipline is another match series, another injury exposure, another allocation of psychological energy.

Second, obligation. WTT imposes mandatory participation rules and financial penalties for withdrawing from entered events. This is the direct friction point between an organiser's commercial interests and an athlete's autonomy over their own body.

Third, points distribution by round. A deep run at a low-tier event can yield fewer points than a quarterfinal at a high-tier one. That makes a player's calendar a optimisation problem, not an emotional decision.

I track the schedules of leading players the way I track an investment portfolio. Each event is a position. Each withdrawal is a position closed, and the question is always the same: why was it closed?

China and the rest of the world

Layer four is the competitive landscape. Here I must say something many people do not want to hear: world table tennis is not a balanced contest. It is a system with one centre and several rings.

The innermost ring is China. That is not up for debate.

The second ring consists of table tennis nations capable of producing a player who can beat a Chinese mainstay in a specific match: Sweden with Moregard and Anton Kallberg; France with the two Lebrun brothers; Brazil with Hugo Calderano; Japan with Harimoto and the next generation behind him; Germany with a long tradition but now in transition as Dimitrij Ovtcharov and Timo Boll reach the late stage of their careers.

The third ring is nations with solid professional foundations but insufficient depth to sustain results over time: South Korea, Chinese Taipei, Hong Kong, Portugal, Slovenia, Egypt.

This picture is stable enough that many treat it as permanent. I do not, but I also do not think it changes within one Olympic cycle.

What mattered at Paris 2026 was that the outer rings began to bite in specific matches, in specific rounds, against specific opponents. Wang Chuqin lost to Moregard in the round of thirty-two. Calderano has repeatedly beaten Chinese players at WTT level. Japan and France keep producing eighteen- and nineteen-year-olds capable of deep runs.

But "can win a match" and "can win a major" are two different levels of the same problem. The difference sits in the third tier of the match — the tier I said appears in no public data table.

Paris 2026 and Truls Moregard were not there to make you believe in miracles; they were there to remind you that probability was never destiny.

From a Korean perspective — the culture I was born into — this gap shows up more clearly than in many other nations. South Korea has a strong table tennis tradition, punishing training, and a training culture closer to China's than any Western nation's. But the gap lies elsewhere. I have written before that the difference is not in basic technique but in three things: how pressure at decisive scores is handled, how placement is chosen when pinned, and how errors are restrained across long games.

None of those three appear on a scoreboard. They only appear when you watch the same rally three times, at three different speeds.

The rulebook is a player too

Layer five is rules and governance. I consider it the most undervalued layer in the entire analytical chain, because it does not act on a single match — it acts on an entire generation's careers.

The most notable recent event in this layer was the withdrawal of Fan Zhendong and Chen Meng from the world ranking, announced on 27 December 2026. Both were singles champions at Paris 2026 — Fan Zhendong beating Truls Moregard in the final, Chen Meng beating Sun Yingsha in the women's final. Two reigning Olympic champions leaving the ranking system at once is a substantial event, and the stated reasons concerned WTT's mandatory participation rules and withdrawal penalties.

My point is not to take a side. My point is the structure of the problem. A ranking system is simultaneously a commercial system. Points are not only a measure of achievement; they are a commodity. They determine entry, seeding position, sponsorship value, and ticket prices for a given session.

When one body organises the events, sells the rights, and manages the ranking, every change to the points rules has winners and losers. The governance layer is not neutral. It is a player.

In my trade, when a new regulation appears, I do three things. First, identify who benefits on points. Second, identify who benefits on calendar. Third, search history for a precedent and check how it played out. Those three steps nearly always predict the parties' reactions before the parties speak.

Governance risk in this sport differs from team sports. In football, a rule change affects twenty clubs. In table tennis, a rule change affects individuals, and every individual is a small business. There is no union, no collective agreement, no negotiation mechanism. That makes governance decisions here more immediate and more direct in their impact than in most team sports.

The pipeline behind the lights

Layer six is coaching staff and the talent pipeline. This layer decides results five years out, which is exactly why it receives the least attention now.

In China, a generational transition is underway. Ma Long — the most decorated table tennis player in Olympic history, with six gold medals — has passed his peak. Fan Zhendong and Chen Meng have left the ranking system. A golden generation departing does not weaken Chinese table tennis immediately, but it changes the team's internal structure: who is the pillar, who is the apprentice, and who carries responsibility for points at team events.

What is interesting is that China does not train by replacing one person with one person. It trains by generating a cohort at once — a generational-skip development strategy that bypasses a middle age band to concentrate resources on very young talents. The upside is more options. The downside is that players who get skipped often leave the system early and never return.

South Korea is different. I have followed Asian youth events for years, and what I see is a country that produces players with excellent fundamentals but lacks a constant competitive environment at the highest level. Shin Yubin is a notable case: she and Lim Jong-hoon won mixed doubles bronze at Paris 2026. That is a signal that South Korea can still produce a player capable of a medal at the biggest event.

But a mixed doubles bronze and a singles medal are different stories about pipeline depth. Doubles allows mutual cover. Singles does not.

Coaching stability is the first variable I check when assessing a national team. A change of head coach disrupts the division of labour, changes doubles pairings, and usually triggers an adaptation period of six to eighteen months. That period appears in no ranking table, but it shows up clearly in results.

A risk surface with no column in the spreadsheet

Layer seven is the risk surface, and I believe it is handled worse here than in almost any other part of the industry.

Table tennis has a distinctive injury profile: shoulder and wrist injuries from continuous high-amplitude rotation; lower back injuries from prolonged crouched posture; knee and ankle injuries from high-intensity lateral movement; and elbow injuries from the recoil force of the rubber. None of these appear in match statistics.

I hold a firm professional position on injury and return to play, formed over years of watching different sports. Rushing back from an anterior cruciate ligament injury destroys the second phase of many athletes' careers. And the psychological fear after injury is harder to repair than the physical damage. A ligament may heal in nine months. The body's natural reflex before an explosive movement takes longer, and no timeline for it is written in any medical file.

In table tennis, the operational risk comes from another source: multi-event load. A player competing in singles, men's doubles, and mixed doubles at one tournament plays three times the matches of a singles-only entrant. Statistical tables record their win rate across all disciplines as a single block. That reading completely hides the reality that their singles win rate may be falling because they spent their energy in the other two events.

This is where I have to say something the forty-seven-page report could not: finding no risks is not the same as there being no risks. Those are different states, and confusing them is the cause of most serious errors in this trade.

The heat of narrative and the cold of data

Layer eight is public narrative and expectation. It does not affect match results, but it affects how everyone reads match results — and eventually how the player reads himself.

Every cycle produces a different narrative label. One cycle is the story of a young player emerging under the expectations of an entire nation. Another is the story of twin stars running in parallel. Another is the story of a dynasty defending its position. And another is the countdown to farewell.

The problem with narrative labels is not that they are wrong. The problem is that they sustain themselves on their own heat, detached from the data underneath. When a narrative gets hot enough, every result is read in its service: a win becomes proof of talent, a loss becomes proof of necessary growth.

I have one strange experience I still use to explain this layer to younger colleagues. It is the period when matches were played without spectators. When the pandemic forced events into empty arenas, I had a chance to observe something normally invisible: the crowd-psychology variable separated from the technical one.

A stadium with no spectators is not an empty stadium — it is a laboratory.

In that laboratory I saw players famous for performing in front of crowds perform better with nobody there. And I saw players considered mentally fragile perform far more steadily. Both phenomena said the same thing: most of what we call nerve is actually the relationship between a player and the atmosphere around them.

In table tennis, where each point lasts seconds and applause happens mostly between points rather than within them, the effect is smaller than in other sports. But it is not zero. And the notable thing is that it is almost never included in any prediction model.

Industry transmission

Layer nine, the last, is industry transmission. It sits farthest from the match, yet it decides matches over the longest horizon.

Table tennis has an unusual industrial structure. The equipment market concentrates around a few major Japanese, German, and Chinese brands, competing mainly in rubber and blades for amateur players. Professional equipment revenue is far smaller than amateur revenue — meaning a small change in a leading player's rubber choice can generate a media effect far larger than its actual effect on results.

On the event side, commercialisation through WTT has increased the number of events, raised broadcast rights value, and professionalised organisation. It has also increased the competition volume each player must absorb to hold position. Those are two faces of one process.

On the labour market side, domestic leagues in Japan and Germany have long been destinations for international players. That mobility creates a shadow transfer market with very little public data. A player's value there depends on ranking, ticket-pulling power, and nationality.

And here I hold a fairly hard position, formed by watching transfer markets across sports. The overvaluation bubble for young players is real, and in some sports it is already deflating. Paying the equivalent of one hundred million euros for a player who has not played fifty top-level matches is a naked wager. In table tennis the mechanism shows up as personal sponsorship contracts for fifteen- and sixteen-year-olds. The sums are far smaller than in football, but the risk structure is identical: paying upfront for an unverified probability.

The notable thing is that the most heavily valued players are usually not those with the highest probability of winning, but those with the most compelling story. Layer nine and layer eight meet there, and the data is not in the middle.

An empty cell is not a safe result

Now I return to the forty-seven-page report.

Having walked the nine layers, it is clear that each one requires a different kind of data, and each kind can be absent silently. Technique needs an equipment name. Player data needs a person. The event system needs an event name and a timestamp. The landscape needs an association. Governance needs a named regulation. The pipeline needs a roster. Risk needs an injury signal. Narrative needs a headline. Industry needs a commercial detail.

When all of those are missing, the result is not a neutral analysis. The result is a document with the shape of analysis and the substance of blank space.

What troubles me most is not the document itself but how it gets read. In most workflows, a document with complete headings and complete sections passes formal checks automatically. Nobody verifies whether the cells contain real content, as long as the cells contain words.

This is the trap I want to name: confusing an empty result with a safe result.

In sports we hit this trap almost every week without noticing.

A player with no injury report is not a healthy player. It means we have no information about their condition. A player who has never lost to a specific opponent will not necessarily win next time. It means the sample is too small to say anything. A national team nobody complains about is not necessarily fine. It means nobody has recorded anything yet.

Correlation is not causation. That is an old line. But there is another, less often said and more important in my trade: the absence of correlation is not the absence of a problem.

In table tennis, the most valuable data is often negative data. The stroke not played. The point not won. The match not entered. The reason for withdrawal never explained. A player reaches the quarterfinal and withdraws, and the news says one word — injury — without saying which injury, where, or how severe.

A good analyst is not the one with the fullest spreadsheet. A good analyst reads the empty spreadsheet, knowing which cell is empty because nothing happened, and which is empty because something happened and nobody recorded it.

The difference between those two kinds of empty cells is my entire profession.

Signals for the Los Angeles cycle

I did not write this to pass judgement on one specific report. I wrote it to say what the next Olympic cycle will demand.

The cycle toward Los Angeles will not reward whoever has the most data. Everyone will have plenty of data. What gets rewarded is the ability to state precisely what your own data does not contain.

I am tracking four signals over the next two years. The first is how associations handle the transition once the current generation of Olympic champions departs — not their results, but whether they publish a clear pathway. The second is the debate over athlete rights within the mandatory participation system, because any change there rewrites the entire calendar and therefore the entire ranking.

The third is the maturity of the cohort born after 2026. Not whether they can beat someone, but whether they can hold a level across three consecutive events. That is the test very few pass at that age, and the test that single-result prediction models always misjudge.

The fourth, and perhaps most important, is whether any association dares to publish an open register of what it does not know. A register of blanks. A list of questions it has no data to answer. If that happens, it will be the biggest step forward in sports analytics in decades — not because it gives us more data, but because it gives us a map of the unmapped regions.

Thirty-six years in this trade taught me one simple thing. We do not gain an edge by knowing more than others. We gain an edge by knowing more clearly than others what we do not know.

That forty-seven-page report, with every cell reading insufficient information, actually told me something more honest than any confident analysis I have ever read. Its only problem was that it did not know it was saying it. And neither did its readers.

A table tennis table is two point seven metres long. The distance between the two sides of the net is one and a half metres. An entire elite career is decided within a space smaller than a living room. And most of what decides it inside that space will never be written into any data table.

My job is to record what is not recorded. The reader's job is to know that an empty cell always has two meanings.