The Source-Data Gap in Martial Arts: What Remains When the Original Disappears

The Source-Data Gap in Martial Arts: What Remains When the Original Disappears

On my desk in Nakano, Tokyo, there is a file that has sat untouched for months. It has proper column headers: article title, core argument, entities, supporting detail, source. Not one cell contains a word. I do not open it to read it. I open it to hold up a mirror, to remind myself that this is the default state of my trade: the skeleton exists, the body does not.

Late last month I sat in the seventh row of a small arena in Tokyo watching a six-bout boxing card. In my hand was a notebook divided into six columns, a habit I have kept for years. The first column held names. The second held results. The remaining four — significant strikes landed, the moment one fighter lost balance, the actual rest interval between rounds, breathing rhythm in the final thirty seconds — stayed entirely blank. I was not lazy. I simply had no source to fill them with.

Two days later I received an extraction sheet with every column header in place and no content beneath any of them. It concluded that no analysis was possible because the base data was missing, and that the label "martial arts" had not yet been split into its two distinct branches: modern combat sports and traditional performance martial arts. I read it and felt recognition. It was my six-column notebook, typed up.

Every sports story begins with a forgotten number. But some stories can only begin once we accept that the number never existed, that it was never recorded, and that its absence is not a technical accident but a structural feature of an entire industry.

That is the subject here: not a particular fight, but the hole beneath every fight.

Why a major sport has no data foundation

Consider a dry comparison. Football has event-data systems tracking thousands of matches per season, each event coded with coordinates and time stamps. Basketball tracks every possession and shot location. Athletics has electronic timing and finish-line imagery resolvable to thousandths of a second. Swimming has touchpad systems and split data for every lap. None of these sports is more magical than martial arts. They simply have something martial arts lacks: a single body powerful enough to force every party to record the same way.

Martial arts is splintered. Amateur boxing has one system; professional boxing has dozens of licensing commissions across dozens of countries. Judo has a relatively centralized international federation. Karate scores through a technical and athletic points system. Taekwondo uses electronic scoring. Wrestling has its own rules. Muay Thai has its own. Mixed martial arts rules are set by individual promotions. And traditional performance martial arts — forms, patterns, choreographed sparring — live in a different world entirely, where "winning" is defined not by landed strikes but by a panel's scale.

The first consequence is decisive: the label "martial arts" is a bag holding things that cannot be measured with one ruler. In an extraction sheet, writing "martial arts" without splitting modern combat sports from traditional forms renders everything downstream meaningless. A wrestler and a forms competitor may carry the same provincial federation card, but their data cannot sit side by side. You cannot compare one person's landed strikes with another person's technical score and call the result analysis.

When an analyst is asked to assess a piece with no source content, refusing is the rational response. The extraction sheet did exactly that. But it also revealed something more frightening: it is not one article missing data. An entire industry is missing data, and that industry has grown comfortable writing without it.

The sheet listed eight analytical layers that become available with base data: technical-tactical, competitive conditions, organizational positioning, commercial value, rules and governance, health risk, narrative value, and industry transmission. It scored all eight at zero, because layer zero — the source layer — was empty.

That is an uncomfortably precise metaphor for how martial arts media has operated for decades. We have every kind of analysis: technical, psychological, historical, cultural. What we lack is the original. The record of what actually happened, with enough parameters for someone else to verify.

Based on my experience following bouts in both Vietnam and Japan, I can say the shortfall takes two very different shapes in the two places, and both are troubling in their own way.

In Japan, a recording culture exists but is fragmented. High school clubs keep ledgers of student results year by year. Prefectural tournaments archive outcomes. But what is kept is usually only the final result: who beat whom, by what score, at what ranking. The process — weight before and after a cut, stoppages for injury, reaction time when pushed into a corner — rarely survives outside a coach's private notebook, and those notebooks are typically thrown away when the coach retires.

In Vietnam, the recording culture exists as oral transmission. Many generations of fighters were raised inside a teacher's memory. The teacher remembers which student was strong in the third round, which one lost composure when behind on points, which one had a habit of resetting stance after the second strike. That data is worth gold and is as fragile as breath. When the teacher dies, an entire database dies with him.

Both models produce the same outcome: to analyze a fight seriously, a writer must rebuild the base data from scratch, every time, for every piece. No accumulation. No shared database. Nothing to cross-check.

The ruleset is the first field left blank

When I read martial arts coverage, the first thing I look for is not the result but the ruleset. Because the phrase "the third round" means entirely different things depending on the sport.

In amateur Olympic-style boxing, a bout is typically three rounds of three minutes. In professional boxing, a championship fight can be twelve rounds of three minutes. International taekwondo competition runs three rounds of two minutes with short breaks. Judo runs four continuous minutes, and if scores are level, moves into unlimited golden score until a decisive point. Wrestling runs two periods of three minutes. Traditional Muay Thai runs five rounds of three minutes. Mixed martial arts can be three five-minute rounds in a regular bout and five five-minute rounds in a title fight.

Which means a sentence like "he faded in the third round" has zero analytical value until you know whether that third round lasted three minutes, two minutes or five, and whether it was the third of four or the third of twelve.

That is only the first layer. The second is versioning. Many federations revise rules on a cycle, sometimes only months after an Olympic Games. The gap between revisions can completely change how a fighter throws and how a coach calculates points. A statistic spanning two rule versions is a contaminated statistic.

There is more. The same sport name under two federations can score differently. The same federation at two events can apply different administrative rules on whether pushing is permitted, whether leg catches are allowed. In that context, collapsing every bout of one athlete into a single index is an anti-scientific act performed unconsciously every day on news desks.

There is a saying I learned from a Japanese judo coach past seventy who had taught in three countries. He said: "Rules are not the frame of the fight. Rules are part of the athlete." What he meant: a judoka raised under one ruleset develops reflexes entirely different from one raised under another, even if both wear the same uniform. Analyze without the ruleset and you are analyzing a person who does not exist.

Measurement and judgment are two different data types

Here I need to state clearly a distinction sports media almost never mentions.

There are two kinds of data in combat sports. The first is measurement data: time, distance, force, speed, target contacts recorded by a device. The second is judgment data: scores produced by humans, carrying the full error of perception, experience, viewing angle, and sometimes crowd pressure.

In sports with electronic scoring, most results belong to the first type. In sports judged by officials seated around the mat, results belong to the second. But on the record sheet, both are written identically: a win. There is no asterisk noting that this victory was the output of a measurement while that one was the output of a vote.

I once spent weeks looking for any boxing federation that publishes an inter-rater agreement index — consistency when multiple judges score the same round. The answer is essentially none. A few scattered academic studies mention it, but no open database lets an outsider verify at scale. Which means the entire history of boxing, at its deepest layer, is built on numbers whose own creators cannot measure their stability.

When everyone looks at the victory, I look for where the weakness is hidden. And the biggest weakness in martial arts is not any fighter's conditioning or technique. It is that the sport's achievement-recording system cannot distinguish a measurement from an opinion.

Weight cutting: the richest and most sealed data zone

If there is one place where martial arts data exists at high density but is almost never published, it is weight cutting.

Official weigh-ins typically occur about a day before competition, sometimes only hours before at smaller events. The data is fully recorded: official weight, weigh-in time, hydration status, sometimes medical notes. But by fight time, a fighter's true weight can be three to seven kilograms above the number on the scale, depending on the sport and division. That is the gap between the published number and the number that decides the fight.

I once sat in a small gym in central Tokyo watching a young fighter prepare for his debut. Over the final two weeks he dropped nearly six kilograms. His coach recorded everything in a notebook: sleep hours, water intake, urination frequency, mood after each session. It was a personal database detailed to an astonishing degree. Not one line was published, archived, or used to analyze anything beyond helping one man make weight on time.

Looking at Vietnam, I see a different model. Many young fighters at provincial level have no structured weight-management process, partly due to thin sports-medicine staffing, partly because dense competition calendars turn weight control into improvisation rather than strategy. The result is that data is not only missing but unreliable when it does appear.

If I wanted to find out why a fighter wins five bouts early in a season and then loses four straight, weight cutting is where I would look first. But to look, I would need something this industry does not offer: access to true fight-night weight, not the number on the scale.

Medicine and injury: data exists but stays locked in a drawer

In many countries, athletic commissions require fighters to pass medical checks before and after bouts, and every case of a health-related suspension is recorded in administrative lists. That is an enormous data source on injury in combat sports: injury type, severity, recovery time, recurrence rate.

But it is almost never aggregated into an open dataset for outside analysis. In most of Asia it sits scattered in paper files across federations, hospitals and event organizers, and disappears when those bodies dissolve or their staff turn over.

Injury in martial arts carries a feature that makes missing data more serious than in other sports: cumulative brain damage does not show up immediately. A strike to the head may leave no trace in that bout's medical record, yet leave a trace ten years later. Research into cumulative trauma has been conducted at multiple centers worldwide for more than a decade, and all of it hits the same wall: control groups too small, historical data too thin, and nobody recorded how many head strikes a fighter absorbed across an amateur career.

When the stadium is empty, the person inside finally speaks. I learned that during the period when global competition shut down. With no crowd, no roar, fighters told me things they had not said before: about mornings when they could not recall the name of a drill, about headaches lasting days after a bout they never reported to a doctor, because reporting meant being struck from the entry list.

That is why I believe the medical data gap in martial arts is not purely technical. It is a question of incentives. Nobody wants that number to exist, because once it exists, someone must act.

The broadcast layer severs the chain of source

There is a form of data loss rarely discussed: the intervention of television itself.

A martial arts bout that reaches air has passed through many hands. The director picks camera angles. An editor trims the walkout, drops the rest periods, speeds through the feeling-out phase. A commentator talks over the referee's count. Graphics overlay statistics on screen — statistics usually recalculated by the producer's own method, with no published methodology.

The result: two people watch the same fight and receive two different chains of facts. One watches the uncut live feed. One watches a four-minute highlight. The second knows the result but not the process, and more importantly does not know anything is missing.

I once had to reconstruct a single round from three different video sources at three different frame rates, purely to establish when one fighter began losing balance. At one speed the collision looks like a major error. At another it looks like a minor accident. The truth is a real physical event; the evidence for it depends entirely on the device and the editor.

This leads to a principle I keep in documentary work: always log the original timecode of every clip I use. Without timecode, a clip is not evidence. It is only a beautiful image that can be cut to serve any intent.

The economics of data: where there is betting, there are numbers

There is a simple rule I have never seen broken in this industry: data follows money, not sporting value.

Commercial mixed martial arts has the most detailed data systems in the entire martial arts family. Figures such as significant strikes landed, takedown success rate, and control time are published for every bout on the promotion's official statistics platform. The reason is not a love of science. The reason is that each of those bouts is a betting market, and a betting market needs numbers to price.

At the other end of the spectrum, traditional performance disciplines have almost no equivalent database. No statistical table tracks a forms competitor across seasons. No index measures the stability of judging panels. No market needs those numbers, so they do not exist.

Put differently: how much a fighter is documented depends not on how good they are, but on whether anyone makes money measuring them. This is one of the great distortions in the entire achievement-recording system of sport, and it explains why a young champion in a Vietnamese province can have thinner records than an amateur at a small Japanese event, despite comparable ability.

Archives are rotting, literally

On a visit to a training center in northern Vietnam, a coach led me into the storage room. It was small and hot, with a cabinet holding nearly twenty old discs recording roughly fifteen years of competition. Many had begun delaminating. There was no other copy. The coach told me he planned to transfer them to a hard drive, but had not found the time, and the center's computer had been broken for two years.

I tell this story not to evoke sentiment. I tell it to show that data loss in martial arts is not abstract. It happens as specific events, with dates, with room temperatures, with dehumidifier packs nobody replaces.

In Japan I saw the same problem in a different form: tapes in obsolete formats sitting in broadcaster and club storerooms, with fewer and fewer devices able to read them. Some programs I wanted to cite for analysis could no longer be played at all. They still exist physically but are functionally dead.

When I say martial arts records are being erased from history, I do not mean it as metaphor.

Vietnam: data living inside the teacher's memory

There is a cultural difference I am always careful about raising, because it slides easily into a cliché about the East.

In Vietnam, a great deal of martial knowledge has never been written down systematically. It passes through observation, through hands-on correction in session after session, through a student following a teacher for ten years and then taking over the gym with memory as the only luggage. This creates a real paradox: coaching quality can be very high while verifiability is close to zero.

A young instructor in Japan is trained inside a system with curricula, grading exams and competency files. A young instructor in Vietnam is trained inside a system where credibility depends on who his teacher was, and later on his students' competitive results. Data about competence exists, but as relationships, not as files.

I believe in data, but I write about what data cannot measure. And one thing data cannot measure is the whole category of knowledge living inside a teacher's body. The day they stop, we lose more than a witness. We lose a database that was never backed up.

The alibi of missing data

Here I have to say something many colleagues will not enjoy.

The shortage of data in martial arts is not purely a tragedy. It is also a comfort.

An industry with full data is forced to face numbers that wound the very legends it sells. When a famous puncher has a lower accuracy rate than an unknown fighter three divisions down, the table forces someone to explain. When a champion club posts a lower performance index than a mid-table side, that number does not vanish. It sits there waiting to be rebutted, and the waiting itself is a form of discipline.

Martial arts has no such discipline. And because it does not, it is permitted to tell stories that cannot be falsified.

There is an instructive counterexample. Taekwondo has among the best data infrastructure in martial arts: electronic scoring, sensors in the chest guard, video review, continuously updated international rankings. If missing data were the only cause of poor analysis, taekwondo should have the strongest analytical culture in the family. It does not. Most taekwondo coverage still stops at results and praise for fighting spirit.

This proves something uncomfortable: data is not the only bottleneck. The remaining bottleneck is the will to publish numbers that do not flatter the story already written.

Breaking convention does not require a loud voice, it requires evidence heavy enough. But heavy evidence must first be allowed to exist. In martial arts, we routinely destroy the evidence before anyone can argue over it.

I also have to guard against myself here. Writers who love paradox tend to turn every gap into a conspiracy and every omission into a cover-up. That is not the case. Most data gaps in martial arts conceal nothing sinister. They are the product of tired people, thin budgets, and a system where no one is tasked with archiving. But precisely because most are accidental, their consequences outlast any deliberate conspiracy.

Five fields that need filling

After years with notebooks, discs, paper files and patchwork spreadsheets, I have a minimal proposal. No world federation required. No hundred-million-dollar data platform. Just five fields recorded for every martial arts bout at every level, including provincial cards.

First, the ruleset with the version applied to that specific bout, stating rounds, round length, and tiebreak provisions.

Second, official weight with the time of weigh-in and, where possible, fight-night weight.

Third, the scoring system used, clearly distinguishing electronic devices from judging panels, with the number of scorers.

Fourth, the list of referees and judges, because judgment data only means something when you know who judged.

Fifth, uncut footage with timecode.

These five fields do not solve everything. They do not measure breathing rhythm in the final thirty seconds, the fear before stepping onto the mat, the moment a fighter understands he has lost before the referee announces it. But they create a hard floor to stand on, and on that floor, every later analysis can be contested — which means it can be improved.

Fans remember the score; I remember the loser's expression in the final second. But to write honestly about that expression, I need to know exactly which bout I am writing about, under which rules, with which instruments of measurement, in whose hands. Without those, emotion is just speculation dressed up as literature.

What I want to leave in the empty file

I still keep that blank document on my desk. I will not delete it, and I will not fill it with guesses.

Over the next decade, martial arts may still have no unified federation, no open database, no index everyone accepts. We may still write about bouts through impression, through a teacher's memory, through the account of someone who stood inside the ring. But each time someone sits down and logs one more field — a weigh-in date, a rule version, a timecode — the hole shrinks a little.

I believe the decency of an industry lies not in what it celebrates but in what it is willing to record. An industry willing to record numbers that work against its own beautiful story is confident enough to last.

If someone opens this file in ten years, I hope they find at least five fields filled in. The rest of the fight, I trust they will write themselves.