The Empty Dossier: The Classification Gap in Martial Arts and the Cost of One Umbrella Label

The timestamp on the file read 21:14. Five fields. None of them contained data. I received that extract on an October evening, after sending forty questions to three different sources about a martial arts event staged in Vietnam. I did not get a draft. I got the final version. The information points field was blank. The entities field was blank. The core viewpoint field was blank. The source quality field was blank. The only thing left intact, bolded on the first line, was the domain label: martial_arts. A file exists. Its content does not. In fourteen years on the job I have read no fewer than two thousand incomplete dossiers. Incomplete because the source died. Incomplete because someone hid something. Incomplete because the reporter was lazy. This file was incomplete in a different way: it was empty at exactly the point where the entire system is empty. No one deleted data. No one edited a file at midnight before the play-off. Nobody could define the thing they were talking about. One October evening, one empty file, one umbrella label. Behind it sits an industry operating on a vague keyword. Martial arts is an administrative label, used as though it were a single discipline. Beneath that label, in Vietnam, at least two entirely different families exist. The first is modern combat sport: boxing, kickboxing, Muay Thai, sanda, grappling, MMA. Their defining traits are direct confrontation, a weigh-in system, round-by-round judging, concussion risk, and an international betting market that tracks every round. The second is traditional and performance martial arts: taolu, forms, weapons, routines scored on a technical deduction sheet. Their defining traits are performance, scoring by error code, and risk concentrated in joints, spine and accumulated training load. These two families share no ruleset. They share no weigh-in system. They share no medical risk model. They do not even share athlete records under the same standard. On paper, they sit under one label. The consequences do not stop at administration. When one label covers both families, everything downstream is dragged off line with it: anti-doping testing pools, sanction frameworks, competition calendars, payroll, sponsorship contracts, coaching licences, injury insurance terms, and the way an integrity monitoring unit decides who needs watching. I tried to trace this label through four different documents from a regional championship: athlete registration forms, weigh-in records, scoring sheets, and the final report. Four documents, four field formats, and not one contained a discipline code distinguishing taolu from sanda. Officials record the event name, not the family. When a computer reads the file, two athletes who differ in physiology, ruleset and risk model become two identical rows. In Vietnam, wushu was long a reliable medal source at SEA Games and ASIAD, in both taolu and sanda. Women's boxing has produced athletes who reached the Olympic stage. Kickboxing and Muay Thai grew hard in the amateur system for over a decade. MMA gyms opened across Hanoi and Ho Chi Minh City, mostly self-run, mostly outside any federation with a registration code. A taolu athlete and an MMA fighter can absolutely end up in the same column of the same spreadsheet, and from that column onward nobody can tell them apart. The umbrella label saves work for the person writing the report. It costs the person who has to make the decision. That night I reopened all my notes on domestic martial arts events since 2026. My method is to cross-check three layers: the entry list, the anti-doping testing list, and the sponsorship contract list. These three layers rarely match. Where they fail to match is where the story lives. Layer one: the laboratory. The laboratory does not know the athlete's name. That is why I trust it. A urine sample is registered by code. The laboratory receives a code, returns a result by code. The chain of trust depends on the collector and the analyst not knowing the athlete's identity. That chain only works if the ordering party knows precisely which kind of martial artist is being tested. The risk profile of a twenty-two-year-old sanda fighter who cuts weight four times in a season is nothing like the risk profile of a taolu athlete of the same age training six hours a day with weapons. One label, two sampling criteria. I have seen two initial samples from the same athlete come back clean. The third urine sample showed what the first two did not dare say. The third sample was collected at a different moment, in a different training block, for a different purpose. The World Anti-Doping Agency's prohibited list is published annually and takes effect on 1 January each year. A list only has force when you know who you are applying it to. A taolu athlete treating a joint injury carries a different medication history from a sanda fighter cutting weight. Pooling them into one testing group dilutes sampling probability and misplaces abnormality thresholds. I cross-checked public data from regional federations over three weeks. No table split testing pools by discipline family. All of them used one word. Layer two: the money. A contract usually runs one page. A dirty contract has an entire annex. In an endorsement deal or a fighter transfer, the most important part is not the figure printed on page one. It is in the annex: weight-class bonuses, penalties for missing a weigh-in, percentage splits on ticket revenue, image rights, and termination clauses on a positive test. I once spent seven weeks matching one such annex against public corporate registration records and found three legal entities with no staff, no assets, and nothing but bank accounts. In the Vietnamese martial arts market, most transactions happen outside federation systems. A fighter signs with an event promoter, the promoter signs with a media company, the media company sells rights to a platform. Four contracts, four entities, and none is required to declare a competition weight class under a standard code. When a dispute arises, the regulator opens the file and sees one word: martial arts. From there, nobody can determine who holds authority. Money does not lie. Money only tells a story if the industry code table is detailed enough to cross-check against. Here, the table is thin. Referees are the most neglected data layer. A boxing referee is trained under ten-point must scoring. A taolu judge is trained on a technical deduction sheet. Two training programmes, two licensing systems, one personnel roster. At local level, the same person can appear in both roles in the same season without a separate competency file for each. When a decision becomes controversial, there is no data to check which standard that person was trained to. The space for subjective judgement inside video review systems is wider than people assume, and the concept of a clear and obvious error is itself a vague clause. In martial arts, where scoring criteria change federation by federation, that vagueness multiplies. Another neglected layer is the youth pathway. A fourteen-year-old trainee at a self-run gym has no medical file, no injury screening record, no monthly weight history. When that athlete turns professional at eighteen, the first four years exist only as a coach's memory. That is a gap no doping test can fill, and it is the same gap the youth transfer market is pricing with ever larger numbers. Layer three: competition integrity. A stadium is spotless. A changing room is not. Combat betting markets run round by round. Each round is an independent market. A small deviation in round seven can be exploited without touching the final result. Taolu has no round-by-round market. Boxing does. MMA does. Muay Thai does. Kickboxing does. Four combat disciplines, four markets, one administrative label. Esports betting is eroding competitive integrity faster than traditional sport because the regulatory framework lags behind. That mechanism repeats in combat sports, only more slowly. New platforms, old framework. An integrity unit wanting to monitor odds anomalies needs to know which discipline a bout belongs to, which ruleset applies, how many rounds, how long each round lasts. Without that data at code level, an algorithm can only flag at a generic level, and generic flags are useless. I once asked three major regional promoters which standard they used to store bout data. All three answered with a different standard. One used internal codes. One used event names. One used fighter names as the primary key. Layer four: data traces. I began an investigation with one odd figure in a payroll sheet. I ended in a room without a number. In a case involving a professional team's payroll, the first discrepancy was a gap between the number of salaried people and the number of names on the competition registration list. That gap led to three subsidiaries, then to an unnumbered room on the second floor of an office building. In the room sat an old desktop computer. On its hard drive was a money allocation table. The trace is not in the file. The trace is in the edit history. Creation date, last modified date, number of opens, last user. The final three lines of that log are usually enough to reconstruct a sequence of events without a single witness statement. A doping record sat on an assistant coach's old hard drive. Last modified: the night before the play-off. With the martial arts label, the situation is worse. When there is no standard data structure at discipline level, there is nothing to cross-check. An empty file cannot be caught in an error. You cannot prove a data field was falsified if that field was never defined as mandatory. You cannot trace money if the industry code of the receiving party is identical to the industry code of the sending party. That is the point I want to sit with longest. An umbrella label turns a technical problem into an exemption zone. Nobody has to hide information for the system to be empty. The system simply needs to have nowhere to put it. Who owns bout data? Promoters own commercial data. Federations own registration data. Anti-doping bodies own sample data. Insurers own injury data. Four owners, four keys, no shared key. When I need to reconstruct an event, I stitch four sources by hand, and the join between them is always a person's name. Based on my experience covering bouts, I once sat three days in an arena watching sanda and taolu qualifiers from the same championship, staged one corridor apart. The sanda side had weigh-in physicians, ice bags, post-bout screening forms. The taolu side had technical score sheets, judges, written protests. Two complete machines. Two different data standards. One organiser. One label on paper. I logged all of it into a spreadsheet. After the championship ended, I tried to look up two medallists from those two events in the same public database. No database contained both. The reverse holds too. When a combat athlete moves abroad to compete professionally, his medical record in Vietnam does not travel with him. The buyer receives one line of health status supplied by the agent. No weigh-in history, no count of knockdowns taken, no post-bout screening forms. That is the data any sports injury insurer needs, and it exists, scattered across paper. There is a reasonable case for the umbrella label, and I have to state it. Merging gives small disciplines a voice. A standalone taolu federation lacks the weight to negotiate entry quotas for a regional championship, lacks the budget to hire an independent testing agency, lacks the capacity to retain a full-time sports physician. Inside a larger label, they inherit infrastructure they would never have alone. Wushu taolu benefits from anti-doping education programmes that self-run traditional gyms have never accessed. That is a real benefit. Over-classification carries its own cost. Splitting one label into twelve separate codes means twelve administrative files, twelve budget lines, twelve potential organisers, and twelve datasets with nobody to run them. Several traditional disciplines in Southeast Asia collapsed after being severed from a larger structure that could pay operating costs. The problem sits elsewhere. An umbrella label at budget and policy level is sensible. An umbrella label at competition data level is a disaster. The two can be separated cleanly: one shared governing point of contact, but a distinct discipline code for each family; one shared fund, but separate testing pools; one shared organiser, but data standards set per ruleset. Both layers were collapsed into a single word, and the layer sacrificed was data. Put another way: this ambiguity is cheaper than clarity. It is a design choice, not an accident. An empty file is an owned blank, and every blank has a signatory. The question I leave with those who sign off: in the discipline code field of next year's record, who fills it in, and which court do they answer to if they get it wrong?