One Label, Three Worlds: Why Combat Sports Reporting Fails at the Classification Stage
**Câu trả lời cốt lõi (≤60 từ):** Nhãn "võ thuật" gộp ba đường ray không thể dùng chung tiêu chí đánh giá: thể thao đối kháng chuyên nghiệp (MMA, quyền Anh, kickboxing), tán thủ và quyền thuật biểu diễn. Khi dữ liệu thể thao gán một nhãn duy nhất cho cả ba, mọi chỉ số phía sau — tỷ lệ knock-out, thành tích, rủi ro cân nặng — đều trở thành lỗi phạm trù. **Dữ kiện chính:** - Ba đường ray võ thuật khác nhau về bản chất: có đối thủ, có dứt điểm, có chấm điểm thi đấu — hoặc không có đối thủ nào cả. - Quyền thuật biểu diễn chấm điểm bằng độ khó kỹ thuật cộng chất lượng thực hiện, không có logic thắng thua hay knock-out. - Tán thủ có đòn dứt điểm nhưng phần lớn trận đỉnh cao được định đoạt bằng tích luỹ điểm, nên tỷ lệ knock-out không phản ánh năng lực thực tế. - Quyền Anh chuyên nghiệp phân mảnh theo bốn tổ chức cấp đai, khiến thành tích giữa các hệ thống không quy đổi trực tiếp. - Tại Olympic Tokyo 2021, một vận động viên nhảy cao đạt 2m35, thất bại ở 2m37 và kết thúc thứ tư — kết quả không xuất hiện trong bất kỳ cấu trúc dữ liệu huy chương nào. **Nguồn và thời điểm:** Báo cáo "Stage-2 Deep Analysis Report — Input Integrity Failure" (tài liệu phân tích nội bộ). Ngày xuất bản gốc không được ghi nhận trong tài liệu nguồn; đây là hạn chế đã được nêu rõ trong chính báo cáo. **Hỏi đáp liên quan:** - Hỏi: Vì sao không thể dùng tỷ lệ knock-out để đánh giá vận động viên quyền thuật biểu diễn? Đáp: Vì quyền thuật biểu diễn không có đối thủ và không có kết thúc trận bằng đòn dứt điểm, nên chỉ số đó không đo được bất cứ điều gì thuộc về môn này. - Hỏi: Rủi ro sức khỏe lớn nhất trong thể thao đối kháng chuyên nghiệp là gì? Đáp: Cắt cân là nguy cơ y tế cấp tính có tần suất cao nhất, phụ thuộc vào cân ngoài mùa giải, cân chính thức, khoảng cách giữa hai lần cân và lịch sử tuột cân. - Hỏi: Một bản phân tích không có dữ liệu có giá trị gì? Đáp: Nó có giá trị chẩn đoán, chỉ ra chính xác điểm gãy ở khâu thu thập dữ liệu hoặc khâu phân loại quá thô.
Two in the morning in Incheon. The ceiling fan hummed like a small engine, and on the screen sat a deconstruction report with every field empty.
Title: N/A. Source: N/A. Article type: unclassified. The information points section had not a single line filled in. The entities section carried a circular instruction — identify from the information points above — while above there was nothing to identify.
Eleven years in this trade have taught me that gaps are more uncomfortable than errors. A wrong fact only surfaces when someone reads closely enough. A blank surfaces immediately, and it always arrives with a question from the desk: what can you write from this?
I almost filled it. I opened a new file, typed the names of a few familiar fighters, and started to reconstruct a bout for which I held zero facts. Then I stopped. That empty report, in itself, is a more worthwhile sports story than most of what I filed this month. It is about the stage readers never see: classification.
When a newsroom attaches the label "martial arts" to a piece, it has just merged three worlds whose rules cannot be exchanged into a single drawer. And every error downstream — the wrong metric, the wrong record reading, the wrong health assessment — begins in that drawer.
One drawer, three worlds
Every newsroom has a stage called deconstruction: pulling title, source, type, information points, entity list and time sensitivity out of a raw document. It is the intake desk for the writer. Without it, the writer has nothing to verify against, only a vague sense that something is happening somewhere.
This time the intake desk returned a shell. Sitting with it, I realised the larger problem sat in the domain label itself: martial arts.
In sports data systems, martial arts is a parent category. Under it sit at least three tracks that cannot share one evaluation framework.
The first track is professional combat sport: MMA, boxing, kickboxing, Muay Thai, submission grappling. Here everything orbits win-loss records, finish rates, stylistic matchups and the economics of pay-per-view events.
The second is sanda, the combat discipline of Chinese wushu. It has competitive scoring, punches, kicks, throws and ring-out situations. But scoring weights throws and caught kicks differently from professional kickboxing, and its event system sits largely within national and continental federation structures.
The third is taolu, the choreographed forms discipline. There is no opponent. No rounds. No knockout. The score is constructed by a judging panel from two components: technical difficulty and quality of execution.
These tracks differ in kind, not in degree. Applying knockout rate to a choreographed form is a category error, the equivalent of scoring a dance competition by goals scored.
Across eleven years covering combat sports for the Korean market, I have watched this error repeat, always in sophisticated form. Nobody writes "the knockout rate of this form is unclear". They write something smoother: "this athlete's competitive record is unconvincing". Behind that sentence still sits the wrong ruler, laid against an object outside its range.
The first casualty: knockout rate
In MMA and boxing, finish rate is a meaningful metric because it measures two things at once: the ability to apply pressure and the level of risk an opponent must accept to step in.
But it only means something when three conditions hold. An identified opponent. Rules that permit a fight to end by strikes. And a competition system that records that outcome consistently.
Sanda satisfies the first and third, but the second is partly blocked: bouts can end by strikes, yet most elite bouts are decided by accumulated points, especially when the gap in level is small. A sanda fighter who wins ten straight may never have made anyone fall. Read that number with an MMA ruler and you will draw an entirely wrong conclusion about that person's actual ability.
Taolu satisfies none of the three. No opponent, no finish, no combat result. The entire analytical vocabulary of combat sport becomes meaningless there.
The difference itself does not trouble me; difference is natural. What troubles me is that our data tables still allow all three objects to occupy the same column.
Records do not translate
A record, on its surface, is a string of numbers that looks the same across every discipline. Wins, losses, draws. But what that string means depends on opponent quality and on the system that produced it.
A 12-1 MMA fighter in a regional promotion does not sit at the same level as a 12-1 fighter in the world's leading organisation. In professional boxing, the practice of building records against carefully selected opposition has been documented for decades; four major sanctioning bodies — the World Boxing Association, the World Boxing Council, the International Boxing Federation and the World Boxing Organization — coexist with separate rankings, so a fighter can be rated highly in one system without having faced the top of another.
In sanda, domestic and international records are two separate worlds in both density and opposition quality. In taolu, a "record" is a sequence of medals tied to named events, not a win ratio at all.
Merge those three into one ranking table and you produce something that looks scientific while being an addition of incompatible units.
Over eleven years I have learned to read a record through a single question: who did this person meet while at their peak? The answer usually tells a different story from the string of numbers. I do not look for the winner in a race; I look for the moment they chose not to quit. That moment does not appear in any record column.
Weight: the biggest risk, and the easiest to miss
In professional combat sport, weight-cutting is the highest-frequency acute medical hazard. It depends on four data points: walk-around weight, official weigh-in weight, the interval between weigh-ins, and a history of missed weight.
Without those four, any assessment of a fighter's condition is guesswork.
What is notable is that the same question — is this fighter safe — has different answers on each track. In MMA and professional boxing, weight-cutting is the central concern, capable of causing kidney failure, acute dehydration and, in extreme cases, death. In sanda, weigh-in procedure and event density differ, so the risk differs in kind. In taolu, weight-cutting barely exists in the professional sense; what exists is another pressure entirely — maintaining a light frame to preserve height and spring in jumps.
A data system that treats all three as one will default to assuming every martial artist carries the same weight-cut risk. Two conclusions then follow, and both are harmful. The first: flagging people who carry no real risk. The second, more dangerous: missing the people genuinely in the risk zone, because their data looks "normal" against a benchmark never designed for them.
Two injuries, two natures
Chronic traumatic encephalopathy is the most serious long-term risk in combat sport. It is tied to the number of head impacts, the number of knockouts, the number of hard-contact sparring sessions and the number of years in the game.
But it exists only where an opponent generates force. Taolu does not produce that risk through the same mechanism. It produces a different cluster: knee and ankle joints under repeated load from thousands of landings, spinal compression in rotational movements, and psychological pressure from having to raise technical difficulty to a level the body is not ready for.
Both are real health risks. Different mechanisms, different prevention, different monitoring.
Here I want to state a professional principle I consider among the most important when writing about athlete health: when a health dimension cannot be verified, the result must not be presented as confirmation that no risk exists. A blank in a health table is a blank, not a tick.
Readers have no obligation to distinguish the two. Writers do.
Who governs whom
A martial arts piece that cannot identify the governing organisation cannot assess anything structural.
Professional combat sport runs on a group of organisations with fundamentally different revenue models. The leading MMA promotion relies on exclusive contracts and a fighter payroll share far below team sports. A large Asian organisation runs several disciplines under one roof — MMA, Muay Thai, kickboxing, submission grappling — creating an ecosystem where the same athlete can compete across two disciplines and two rankings. Professional boxing fragments across four title bodies. Kickboxing has separate European and Japanese systems. Submission grappling runs a biennial tournament format. Sanda and taolu sit under a dedicated international federation with its own world and continental championships.
Assigning the wrong governing body cascades. Contract terms are misread. Revenue is misestimated. An athlete's freedom of movement is misjudged. And most importantly, the athlete's position in the system is misread — whether they sit at the top of a small system or in the middle of a large one.
The Korean context makes this vivid. Taekwondo holds strong institutional standing through the education system and Olympic success. MMA arrived later but produces stars with far greater media reach than the market size suggests. Domestic professional boxing has contracted for decades. Three systems, three ways of counting a record, three pay structures, three development pipelines. Writing about them with a single vocabulary is the fastest route to being wrong.
The empty report and the temptation to fill
Back to that screen at two in the morning.
What an empty report taught me is this: most risk in this profession does not come from missing information. It comes from the writer's inability to tolerate missing information.
The gap between "I have not found the data" and "I will write a sentence that sounds plausible" is very thin. Many articles have stepped across it unnoticed, including by the person who wrote them.
In combat sports that gap is twice as dangerous. A wrong tactical read in football gets corrected within hours by someone rewatching the tape. A wrong read on a fighter can survive for years, because most readers have no way to verify — they do not know what that person trained for six months, which injury has not healed, which contract is under negotiation.
An empty analysis, handled correctly, carries more diagnostic value than a full one that is wrong. It points precisely at the fracture: the data acquisition stage failed, or the classification stage applied too coarse a label, or both.
It also reveals something about the structure of the industry. When a system finds no information, it tends to produce information. That is a technical property, and also a human one under pressure to ship.
What the industry has backwards
The prevailing belief is that more data means fewer errors.
I think the opposite is happening.
When the volume of data grows faster than the capacity to classify it, the error rate does not fall — it rises, and the errors simply become harder to detect. A reporter without numbers used to know they had no numbers. A reporter with ten dashboards can believe they understand ten things, when in fact they are reading one thing split into ten columns.
The industry's real shortage is not data. It is taxonomy.
A concrete test for anyone writing about martial arts: can the data you hold answer the question you are asking? If not, the correct move is to change the question, not to force the data to answer.
There is another side to this, and it is the part the industry discusses least. When our classification is too broad, we do not merely lose precision. We lose people.
Tokyo, 2026, in a stadium with no spectators. A high jumper clears 2.35 metres, fails at 2.37, and finishes fourth. No medal. No camera records the moment he steps down off the runway. The coaches' applause rings out thinly in empty stands. Tokyo gave him no medal, but it gave me a story I will tell for as long as I write.
That fourth place does not appear in any data structure I know. It is not a record, not a medal, not a rating point. Yet it is the truest material that sport produced on a silent evening.
I have never run a step in my life, and yet I understand how to touch a finish line with my heart.
That is why I distrust the optimism about data more than I distrust data itself. Empathy and accuracy are not opposites. Both simply need to be placed correctly. Empathy for people; coldness toward systems, tables and labels.
And in combat sports, where a small classification error can turn a sanda fighter into a low-achieving MMA fighter, or a forms competitor into someone with a weak competitive mentality, classification is not a back-office issue. It is the writer's job.
Closing
Every sporting record is a reminder: a limit is only a whisper that has not been heard yet.
But to hear that whisper, you have to know where you are standing. A form is not scored by knockouts. A sanda bout is not read through a professional boxer's ledger. A fighter on a three-loss streak does not automatically become a data point on decline, if the writer does not know who those three opponents were.
That night in Incheon, I chose not to fill the blank. I sent the empty shell back to the intake desk with one line: the domain label is too broad, it must be subdivided before analysis.
That is the lesson from a night with no news.
Sport is a dream that has been quantified; I am only the one recording it in words. And sometimes the most honest thing a chronicler can do is record this: here, there is nothing yet to record.
