Nine Dimensions and an Empty Sheet: When Analysis Must Learn to Stay Silent

Nine Dimensions and an Empty Sheet: When Analysis Must Learn to Stay Silent

Opening: two kilobytes and a decision at 2 a.m.

Tokyo, 2:14 a.m. I open the file operations pushed through, named with a meaningless string and a v3 suffix, two kilobytes in size. Three hundred header rows, not one content row. The information-points column is empty. The entities column is empty. The core-viewpoints column is empty. The source-quality column reads exactly one phrase: not yet assessed.

Twelve years in this trade, I am used to unfinished reports. But this was the first time I received a framework complete in form and absolutely empty in substance. Nine analytical dimensions. Nine tables. Nine conclusion blocks. All formatted correctly, numbered correctly, bolded in the right places. And all carrying the same sentence: insufficient information to assess.

The night-shift colleague before me sent one short line: "Do you want me to make up a few names so it looks presentable?" I answered four minutes later, slower than a refusal needs. Because I understood the pressure behind the question. An empty table sent upstairs looks like a failure. A table with names, numbers and conclusions looks like a productive day. The distance between those two things, in sports analytics, is usually a few minutes of invention.

I declined. This article is why.

Context: why a nine-dimension framework exists at all

Athletics is the strangest sport in the professional ecosystem. It is the only one whose final result needs no argument. No disputed goal, no contested corner, no four-minute video review tracing a toe. There is a number, and that number is correct or incorrect under measurement rules.

Precisely because of that clarity, athletics is the sport most often analysed badly.

Viewers see the final number and believe they have understood the whole story. One athlete runs 100 metres in 9.86 seconds. Another runs 9.92. The conclusion arrives instantly: the first is better. But that 9.86 was measured in what wind, at what altitude, in what shoe, in what week of the training cycle, after how many recovery days, before or after a transcontinental flight. Six questions. And among those six answers, three could reverse the original conclusion.

That is why the nine-dimension framework exists. It was built not to complicate a simple sport, but to resist the most natural human instinct: concluding too quickly from too little.

The nine dimensions are: event and performance; athlete condition; competition structure and qualification mechanism; event landscape and national comparison; rules and anti-doping; team and training system; risk landscape; public narrative and expectation; and finally transmission into the sports industry.

Based on my experience watching thousands of athletics and swimming competitions over twelve years, I can put it briefly: these nine dimensions are not nine independent questions. They are nine defensive layers. Each exists to catch a specific kind of error the previous layer missed. Skip one, and the conclusion leaks at exactly that layer.

And that night in Tokyo, all nine layers were empty.

Dimension one: event and performance, where numbers do not speak for themselves

The first layer sounds simplest. There is a mark, compare it with the record, the qualifying standard, the contemporaries. Done.

But this layer holds five traps, and all five are so common they have become professional habit.

The first trap is wind. World Athletics rules recognise a sprint or horizontal-jump performance as a record only when tailwind does not exceed 2.0 metres per second. A 2.1 m/s gust turns a record into a number that does not exist on paper, even though on the track the athlete ran the same distance before the same crowd.

The second trap is altitude. Thinner air in cities above roughly 1,500 metres reduces drag, and the benefit differs by event. A 200-metre world record set in a high-altitude stadium and one set in a coastal city should not sit side by side without a footnote.

The third trap is equipment. Since 2026, the wave of carbon-plated, ultra-light-foam shoes has shifted the baseline in almost every distance event. An athlete running a marathon in 2026 in the same time as one in 2026 is living in a different sport. Comparing without adjusting for equipment is comparing different units.

The fourth trap is small samples. One beautiful mark in one competition is one data point. One data point is not a trend. The industry has a cruel saying: every athlete has one good day, and every bad analyst builds a career on it.

The fifth trap is unratified numbers. Training runs, internal time trials, hand-timed sprints, exhibition runs with pace cars and scripted hydration. These numbers exist to publicise, not to compare.

In my two-kilobyte file, layer one contained nothing at all. No mark, no event, no measurement conditions. There is no way to assess a performance that was never written down.

When data speaks, laughter is only noise. But when data has not yet spoken, silence is the only honest answer.

Dimension two: athlete condition and the curve nobody sees

The second layer asks about the human behind the number. Four axes must be plotted together: the year-by-year personal-best curve, current-season form, injury risk, and peaking strategy.

The personal-best curve is the most undervalued tool in the industry. A 200-metre runner with a personal best of 20.40 at 19 and 20.15 at 26 is on an entirely different trajectory from an athlete who ran 20.15 at 19 and then stood still for seven years. Both have the same personal best. Two completely different stories.

The age curve differs by event group. Sprint and jump events typically peak earlier, commonly between 24 and 27. Distance events, especially the marathon, peak much later, usually between 28 and 32, and occasionally beyond. Applying one age chart to both groups is the most basic error a young analyst can make.

Injury risk is the axis where public data is weakest. Federations publish start lists, not medical files. Analysts must reconstruct from fragments: withdrawal counts, gaps between consecutive competitions, coaching changes, training-base changes. That is inference from traces, closer to archaeology than statistics.

Peaking strategy is the hardest axis, because it requires internal information. One athlete may be in a loading phase, and a modest result in a minor May meeting is a good sign for August. Another may have peaked in June, and a beautiful string of results is flowing toward a cliff.

In my file, there was no athlete name. No curve. No age, no personal best, no competition calendar. I could not even tell whether the event was running, jumping, throwing or walking. A form table without an athlete is a table that does not exist.

Dimension three: competition structure and qualification mechanism

The third layer is where technical analysis meets administrative law, and where many good forecasts are destroyed by a clause nobody read.

There are three routes into a major championship. The first is meeting a performance standard inside the qualifying window. The second is accumulating world-ranking points. The third is national selection through a quota or domestic performance.

These routes do not fully substitute for one another. An athlete can hit the standard but lack ranking points from too few competitions, or hold points without the standard, or hold both and be removed by a national federation quota. The official world ranking system, in operation since 2026, made this picture considerably more complex, because it turned the quality of minor meetings into part of the equation.

That produces a strategic paradox. Sometimes the optimal action for an athlete is not to compete. Skipping a meeting to protect the body for a bigger one, accepting a points loss, is an investment decision rather than weakness. But it is only optimal if the points model is computed correctly.

Another rarely discussed risk is schedule density. An athlete chasing a standard in the final weeks of the window often competes heavily, travels constantly, sleeps little, and arrives at the main event depleted. The ticket can be earned. The form to use that ticket has already been spent on the way.

And athletics has a variable football does not: relays. An athlete may have to choose between an individual event and a team event when the schedules overlap. That decision is not in the athlete's hands. It happens in a meeting no analyst attends.

There was no competition tier in my file. No deadline, no qualifying standard, no time window. The third defensive layer was entirely absent.

Dimension four: event landscape and the map of national strength

The fourth layer draws a map. Who dominates, who is rising, who is falling behind, and how wide the gaps are.

Athletics has structural geography. Long-distance events have been shaped for decades by a cluster of East African nations, with high-altitude training camps at the centre of that ecosystem. Men's sprinting has had a Caribbean hub and a North American hub over the past two decades. Jumps and throws are more widely distributed, tied closely to school and club systems.

But the map does not stand still. Three shift signals matter every cycle. First, generational gaps in traditional powers: when a golden generation retires within two or three years, the void usually takes one Olympic cycle to fill. Second, the emergence of nations newly investing in youth development. Third, the movement of naturalised athletes and nationality switches, which can change a specific event very quickly.

A common mistake when reading the map is counting medals instead of measuring depth. A nation with one champion and nobody in the final is a thin nation. A nation with five semi-finalists and no champion is a deep nation. Over four years, depth wins. On one night, thinness can win through one outstanding individual.

My file held no country, no athlete, no event. No map. No axes.

Dimension five: rules and anti-doping, the most neglected defensive layer

This is the layer analysts push to the end of the report, and that is a mistake capable of destroying an entire model.

Anti-doping is not a moral topic. It is an operational variable. The out-of-competition testing system requires athletes in the testing pool to file whereabouts information quarterly, and multiple missed tests can lead to a sanction equivalent to an adverse finding. That means an athlete can lose eligibility over personal scheduling, not over any prohibited substance.

There are areas where rules intervene directly in biology and therefore in results. Regulations on testosterone levels and eligibility for certain women's events in athletes with differences in sex development have gone through several revisions, several legal challenges, and remain contested. Since Russian athletes were restricted in certain periods, the concept of the authorised neutral athlete appeared and changed how national totals are counted in ranking tables.

There are also small but decisive details: sole thickness limits, spike counts, carbon-plate elasticity, starting-block geometry, implement diameter and weight. Every small change creates a new baseline, and every prior mark must be reread inside the new frame.

The consequence for analysts is clear: an athlete with an unresolved legal exposure cannot enter a long-horizon model at the same weight as a fully clean one. Not as a judgement. As a probability of appearing.

My file had no athletes, so no legal risk, no precedent, no sanction scenario to project. The fifth layer was empty in the most comfortable way: nobody to suspect.

Dimension six: team and training system

No athlete wins alone. But no athlete wins because of a collective either, and athletics sits at a strange equilibrium between the two.

A modern training group includes a head coach, a strength coach, a biomechanist, a physician, a nutritionist, a psychologist, and a competition manager. Each role can be a strength or a fracture point. A coach excellent at volume but weak at speed work suits a distance runner and harms a sprinter.

The training model is also a geopolitical variable. There is the state-centralised model. The professional club model tied to a market. The university model, common in some school systems. And the far-from-home model, where an athlete lives for years in another country, trains in an international group, and competes under their own flag.

One warning sign I track personally: a coaching change at the wrong time. If an athlete switches training groups within six months of a major championship, all their historical data loses predictive value. The old model describes an old person.

My file had no coach, no group, no model. Nothing to assess for stability.

Dimension seven: the risk landscape, where the first risk is the data

In the standard framework, risk is grouped into categories: competitive, doping, financial and career, rules and eligibility, media and brand, and systemic.

Each has a probability-impact matrix and mitigation measures. That is a correct and necessary practice.

But that night in Tokyo, the matrix collapsed into a single line, and that line sat at the top, upstream of every other risk: data-integrity risk.

The biggest risk in sports analytics is not bad data. It is a table that looks perfect but is empty, because it never triggers anyone's instinct to doubt.

Bad data has noise. It has an outlier, a wrong unit, a misplaced decimal point. The reader frowns. A well-formatted empty table has no noise at all. It just looks like an administrative document. And nobody reads administrative documents closely.

That is the paradox of this whole story. A failure with sound gets fixed. A silent failure gets forwarded.

Dimension eight: public narrative and the expectation gap

The eighth layer measures the temperature of the story around an athlete or event.

Every sports narrative has a life cycle. It is born from a result, amplified by media, priced by markets, and then tested by reality. The analyst's question is not whether the story is interesting, but how much of it rests on real foundations.

There are familiar narrative labels: emerging phenomenon, record assault, comeback, farewell, and the worst of all, doping controversy. Each has a different life cycle and a different durability.

The emerging-phenomenon label has the shortest life. It is usually built on a sample of one to three competitions and collapses fastest when the calendar thickens. The record-assault label lasts longer, because it has a concrete time target. The farewell label is almost always mispriced in both directions, because public emotion and actual capacity move in opposite directions.

The expectation gap is the central measurement tool here. Market, media and public expectations are placed beside the objective assessment from the previous seven layers. The distance between them is where hidden value sits.

I do not guess football; I measure the distance between expectation and goals. In athletics, I measure the distance between expectation and the track.

My file had no narrative, no commentary, no sentiment indicators, no odds. No temperature to read.

## Dimension nine: transmission into the industry The final layer extends beyond the track.

A peak performance transmits through many channels. It lifts the commercial value of a meet, raises broadcast rights, reshapes schedules to serve larger markets. It pressures equipment manufacturers, producing new shoe generations and technical-limit arguments. It opens individual sponsorship contracts and changes the income of a whole generation of young athletes behind it, because local academies and federations see a viable career path.

It also transmits down the youth talent chain, the channel with the longest delay and the deepest effect. One major-championship medal can visibly raise youth enrolment in a country two to three years later. That cycle is usually longer than the tenure of a national head coach.

An analyst working this layer must accept that conclusions cannot be verified in the short term. That is why the layer is often dropped from fast reports. But dropping it means dropping the long-horizon part of the picture, and the long horizon is the only part that can produce a real information edge.

My file had no brands, no products, no markets, no athletes. No flow to draw.

The contrarian angle: the temptation of a complete table

By now the story is clear. One nine-dimension framework, nine defensive layers, and a two-kilobyte file with nothing in it.

The normal professional reflex is to fill it. That reflex is trained, not malicious. When a table has headers, the human brain wants to complete it. That is how we were taught at school: a question has blanks to answer, not to cross out.

But there is a professional truth I learned after years, and it runs against that reflex.

Every mockery is a data column that has not yet been labelled. And every gap filled by speculation is an error that has been labelled wrong.

In 2026, when I was a second-year student in Tokyo writing a data-driven blog, an online commentator dismissed me outright. I had shown that sprint-volume and pressure indicators from a weaker team could reverse the result once the stronger team had nothing left to play for. That night the result matched my calculation. But what I remember is not the joy of winning. What I remember is the fear right before kick-off.

I was afraid because I knew that if the result went the other way, I would have nothing to say. Nothing at all. My model had only two outcomes: correct, or nothing. A model without a grey zone is a model that has not been tested enough.

Three years later, in a meeting room at a Tokyo betting-analytics firm, I presented on a team's off-ball pressure index before a major final. The index measures how many passes an opponent is allowed before being disrupted. The team I analysed was clearly lower than its opponent. That index never kicked a ball that night, but it carried a team to trophy night. A colleague laughed and said machines cannot understand the psychological pressure of a packed stadium.

I did not throw a chart at him. I projected the last thirty matches and said one line: in the meeting room, emotion asks, data answers. That argument ended with a new contract for my group. But what I kept from that night was not the victory. It was the realisation that my colleague's emotion was also data. It was not wrong. It was simply not placed correctly in the model.

That is why I did not fill the empty sheet.

Because there is a kind of error worse than reaching no conclusion. It is reaching the right conclusion for the wrong reason, and then turning that into method.

If I had invented three athlete names that night, assigned them three plausible marks, drawn three convincing age curves, and written a balanced conclusion, my report would have been praised. It would have gone upstairs. It would have been used as the basis for some resource-allocation decision.

And it would have been entirely wrong.

The danger is not the three invented names. The danger is that the table becomes precedent. Next time another empty file arrives, the night-shift analyst remembers that invention was fine last time. Three times, invention becomes procedure. Ten times, procedure becomes culture. And a culture of invention in sports analytics does not produce small errors. It produces wrong investment decisions at scale, and young athletes misjudged for years.

There is a second contrarian point worth stating, even if it is unpopular.

Sports analytics tends to worship completeness. A long report is a good report. A table with many indicators is a serious table. But in many cases, a three-hundred-page report has two hundred and ninety pages of data collected because it could be collected, not because it answered any question.

Information value is not proportional to data volume. It is proportional to questions answered divided by questions asked.

And sometimes the correct answer is: not answerable yet.

I know this sounds like evasion. In twelve years I have been called evasive no fewer than ten times. But I distinguish two very different things. Evasion is when the data exists and you dare not conclude. Honesty is when the data does not exist and you say so.

People routinely confuse the two. And when they do, they reward the second by calling it courage.

What remains: the signal of the next cycle

Morning came to Tokyo as I closed the file. Three hundred header rows, not one content row.

I sent back a report of nine sections, each stating clearly that there was insufficient information to assess, with the first condition in the risk section: send the original file. No invented names. No fake curves. No balanced conclusion at the end to please the reader.

It was not a good report. But it was a correct one. And in this trade, that order matters.

I do not know which athlete will win the next cycle. Nobody does. Anyone who claims to know is selling you something else, not data.

But there is one signal I will track in the coming cycle, and it is not on the track.

It is in the data pipelines. In files that arrive late, in columns left blank by a connection error, in tables finished at the last minute before a meeting.

Because that is where the largest errors are born, and where the fewest people look.

When data speaks, laughter is only noise. But when data is silent, a professional must have the courage to be silent with it.

The question for the next cycle is not who will run fastest.

It is: how many empty tables are being filled right now, somewhere, by people who believe a complete table matters more than an incomplete truth?