Two Bodies in One Data Column: Vietnam's Martial Arts Injury Classification Problem

In May 2026, inside the medical logbook of a provincial sports training centre, a single line was written: Male, 19, martial arts, ankle sprain, 10 days off. Two months later, another line: Female, 21, martial arts, metacarpal fracture, 6 weeks off. The same word, martial arts, sat in the classification column. The same person wrote it. The same book held it.

Behind those two lines are two bodies that are difficult to compare. The first case is a forms athlete, injured at the moment of landing after a spinning jump. Nobody touched her. The second case is a combat athlete, injured by a kick that landed. Somebody touched her, and that contact is the entire point of the sport.

Two mechanisms. Two load-bearing structures. Two recovery protocols. Two timelines for return to competition. One data column.

The crack in the Buriram data was never an error; it was a door. In Vietnam, that door is opening at the narrowest point of the whole machine: classification. Not in the operating theatre, not in the gym, but in the first line of the medical form.

Context: many systems, very few data fields

Vietnamese martial arts is not one block. It is a cluster of parallel systems, each with its own federation, competition calendar, rules and training culture. Vovinam has a belt system with both forms and combat content. Wushu splits into two branches that are almost two different sports: taolu for forms with bare hands and weapons, and the combat branch. Boxing has an amateur side inside the national sports system and a professional side operating on contracts. Muay, kickboxing and pencak silat belong to the direct-contact group. Traditional martial arts are a universe of their own with hundreds of schools, mostly outside the high-performance system. Since 2026 Vietnam has also had a domestic professional MMA stage, LION Championship, reported in local media as the country's first professional MMA promotion, adding a new group of athletes to the injury picture.

On the medical side, the system includes national sports training centres in Hanoi, Da Nang, Ho Chi Minh City and Can Tho; hospitals with sports medicine departments; and local health networks where young athletes train daily. Competitive pressure arrives in cycles: SEA Games, ASIAD, Olympic qualifiers, national championships, youth championships and invited international events.

Three names frequently appear in the same sentence about Vietnamese martial arts, and they represent three entirely different bodies. Duong Thuy Vi, a wushu taolu athlete with multiple SEA Games gold medals in bare-hand and weapon forms. Nguyen Thi Tam, a female boxer who competed at the Tokyo 2026 Olympics. Nguyen Tran Duy Nhat, the most widely recognised Muay Thai fighter in Vietnam, who has competed in ONE Championship. Three sports, three injury profiles, three training cycles. In most current summary reports, all three sit on one line.

SEA Games 31, hosted by Vietnam in 2026, shows how dense the calendar can be. The host nation topped the medal table with 205 gold medals according to figures published after the Games, with martial arts contributing significantly. A successful Games creates an effect few discuss: reporting pressure. When results rise, organisations need to present figures on squad size, training volume and injury situation. But when the storage system lacks sufficiently fine classification fields, people are forced to aggregate. Aggregation is where the truth begins to blur.

My point is not a shortage of equipment or doctors. It is that the data is written with one word, martial arts, for everything. I trust quietly archived numbers more than loud promises, but a number is only trustworthy while it still holds the shape of what it describes.

Three minimum classification axes

In working with sports injury data, I have settled on one principle: every injury needs to be described along at least three axes before it is aggregated. Miss one axis and the aggregate will be wrong in a way that cannot be fixed by adding more data.

The mechanism axis separates contact from non-contact. Combat sport is direct, repeated contact; every sparring round is a chain of controlled but not fully controllable impacts. Taolu is non-contact. Taolu injuries come from repeated load: hundreds of landings in a single session, thousands of knee flexions at the same angle, and rotational movements that force the lumbar spine to work at maximum range. Both are called martial arts, but one is a problem of external force and the other a problem of internal force and repetition.

The load-bearing axis shows which part of the body is carrying the stress. In taolu, it is the ankle, knee, lumbar spine, Achilles tendon and foot, the structures that absorb landing force. In combat sport, it is the hand and wrist from punching and from the returning force, the ribs and flank from kicks and knees, the jaw and facial region, the shoulder, and the soft tissue around the eye. In MMA, add the neck and knee joints during grappling, plus accumulated head impact risk. A single knee-injury percentage calculated across both groups conceals an important difference: knee injuries in taolu are largely chronic overuse injuries, while in combat sport they are largely acute injuries from sudden torsional load.

The workload axis measures the volume the body carried before the injury. In taolu, the measure is landings per session and repetitions of the routine. In combat sport, the measure is sparring rounds per week, fights in the last 30 days and the weight-cut cycle. These three measures are not interchangeable. No formula converts a forms session into a sparring session.

These axes are not academic. They are the minimum conditions for a coach, a doctor or an administrator to make a correct decision.

When aggregated, the average becomes irresponsible

I keep a dataset I coded by hand from competitions and sessions I have observed directly in the region, covering 214 injuries recorded with enough detail to classify. This is a personal dataset, not official figures from any federation, and I say so every time I cite it.

Aggregate everything into one martial arts group and the average return-to-competition time lands around 31 days. That number sounds very reasonable. It is also useless.

Split by branch and the picture changes. The taolu group with mild ankle and foot injuries averages around 18 to 22 days back, mostly grade I and II sprains that recover quickly without complications. The combat group with hand and wrist injuries takes far longer, commonly 45 to 60 days, because the problem is not only bone healing but rebuilding force perception and the tolerance for the returning force of a punch. The knee ligament group in combat sport falls between 7 and 9 months, sometimes longer for athletes with a previous injury at the same site.

Three timelines: 20 days, 50 days, 240 days. Their average is 31 days. Nobody in those three groups recovers in 31 days. The average describes no one, yet it is the number most frequently used in meetings.

The body never negotiates; it only quietly signs the sentence in advance. An aggregated percentage does not make a body slower or faster. It only moves responsibility somewhere else.

Weight cutting: the forgotten variable in every report

There is a data field almost nobody records, and it changes how every combat-sport injury should be read: weight differential in the 72 hours before competition.

Weight cutting is a reality of boxing, Muay and MMA. An athlete can drop several percent of body mass in two days by restricting fluids, training in heavy clothing and eating plain food. The body is then rehydrated in the window between weigh-in and entering the ring. If that window is short, the athlete starts the bout with circulating volume not yet restored, reduced tissue elasticity and slower reflexes.

In the combat group of my coded dataset, injuries occurring in the final rounds are markedly overrepresented compared with early rounds, even though early rounds account for more accumulated minutes. Without a weight-cut field, that pattern is attributed to poor conditioning or to bad luck. With the field, it becomes a cycle management problem.

This is why aggregation harms in a way few notice. Taolu has no weight cutting. Traditional martial arts have no weight cutting. If weight-cut-related injuries are mixed with taolu overuse injuries, both rates are wrong and neither group receives the right solution.

Head injury: where aggregation crosses an ethical line

If I could pick only one reason to separate martial arts data into two storage systems, I would pick head injury.

Taolu has no head contact. Traditional martial arts mostly do not either. Combat sport does, at a measurable frequency. Every sparring session, every round, every bout involves a certain number of impacts to the head, most of them below the threshold that produces clear symptoms, but accumulating month after month and year after year.

Record everything in one column and the head-injury rate across all martial arts becomes suspiciously low. That low number is then used to argue the sport is safe. What is being measured is not the risk of combat sport; it is the average risk of a group diluted by disciplines with no head contact at all.

For head injury, the data must be thicker, not thinner. Head contacts per week. Impacts hard enough to stop training. Full rest days after being knocked down. These three fields cannot be replaced by any composite index, and they belong in a separate place, not mixed into a spreadsheet shared by forms and combat alike.

Competition surface: a data field that barely exists

There is one more variable I have tracked for years and almost never see in any Vietnamese report: training and competition surface.

Taolu trains and competes on matted flooring. Boxing competes on a padded ring canvas, but trains on harder floors. Muay holds outdoor events on harder surfaces at higher temperatures. MMA competes inside a cage, whose mat is a different kind of hard surface, and the cage itself produces a distinct injury type when athletes are driven into the mesh.

Each surface generates its own injury profile. On soft mats, the ankle absorbs more rotational load. On hard floors, the heel and knee joint absorb greater impact. In the cage, contact with the mesh produces hand and wrist injuries in unfavourable positions.

Based on my experience watching bouts at domestic and regional events, I keep seeing one repeating pattern: when the same athlete moves from matted flooring to a hard floor, foot and heel injuries rise over the first two weeks, then decline as the body adapts. That kind of information can be captured in a single box on a medical form and used to design a surface transition plan before every tournament.

Amateur and professional: two sports wearing one name

An amateur boxer competes three rounds of three minutes. A professional boxer competes ten to twelve rounds. A Muay bout runs five rounds. An MMA bout runs three five-minute rounds, or five five-minute rounds in title fights. Round count is not merely duration; it is the number of times an athlete must cycle through recovery between rounds, the number of decisions made while tired, the number of times a fighter accepts damage to gain position.

Aggregating these three groups into one column means the injury rate of a twelve-round professional is calculated together with that of a three-round amateur. The high-risk group is pulled down, the low-risk group is pulled up, and neither receives appropriate measures.

I have declined three television invitations on this subject over the past two years. Not because I do not want to speak. Because on air I would be asked for one single number for Vietnamese martial arts, and any single number I gave would become a false certainty quoted for years.

Traditional martial arts: the largest blind spot

Vietnamese traditional martial arts encompass hundreds of schools and a practitioner population no agency can count precisely. It is the least injury-surveilled group, and also the group with the least access to sports medicine. People train in local schools, community halls and privately run clubs. When injured, they self-treat, rest, or keep training.

From the perspective of a data system, this is a blind spot. From the perspective of a body, it is real risk. The two are routinely separated in high-performance discussions, which care only about the group capable of winning medals.

I hold no illusion that one article fills that blind spot. But I know one thing: when a group has no numbers, that group does not exist in resource allocation decisions. The quietest summer is when I take the most notes, and the quietest season of Vietnamese martial arts sits in the training halls, not in the national centres.

The counterintuitive angle: more data is not automatically better

The easy conclusion is that Vietnam lacks data. I think the more precise problem is that the data has the wrong shape.

If an organisation starts collecting more while keeping a single martial arts column, the outcome will not be better understanding but greater misplaced confidence. Reports will be thicker, charts prettier, and conclusions more wrong, because they will be presented with the certainty of a large dataset.

Another point, counterintuitive in a different way: responsibility does not sit with individual coaches. I have read many analyses blaming a specific coach's overloaded programme. Most of the cases I re-checked traced back to the competition calendar, travel distances between events, the absence of a designed deload cycle, and the absence of a data field capable of early detection. Blaming one person is the fastest way for a system to avoid changing.

One more point concerns taolu. Because it has no contact, it is usually filed as low risk and low medical priority. But repeated-load injuries in taolu behave differently: they arrive slowly, accumulate over years, and can follow an athlete through an entire career as chronic back and knee pain. An injury that never sends anyone to hospital can still end a career. Treating it lightly is a resource-allocation error, not a medical judgement.

And there is one professional habit I have kept for years: every number I cite passes three verifications. First against the original source. Second against an independent source. Third against my own dataset, to see whether the number holds when placed beside other cases in the same group. Three passes, then publication. Not to appear careful, but because in this field a wrong number outlives the person who wrote it.

What it takes to start

If I could propose one minimum change to every martial arts injury form, I would propose five fields. A discipline field, recorded specifically and separating combat from forms. A mechanism field, stating direct contact or non-contact. An expected days-off field. A field for sparring volume and bouts in the preceding 30 days. And a field for the weight differential in the 72 hours before competition, where applicable.

These five fields require no new equipment, no new software, no additional staff. They require one decision: accepting that two different bodies cannot share one column.

Once those fields exist, every summary report afterwards begins to say something. And if you are holding a martial arts injury dataset anywhere, in a medical room, a training hall, a club, I would like to know how many columns it is written in. Buriram taught me that data is never innocent; it is only waiting for a reader.