Empty Profiles in the Transfer Window: When Combat Sports Data Cannot Classify a Human Being

In a small apartment in Busan, I typed a name into three different databases. The name belonged to a fighter I had once watched compete in a provincial arena, a hall with two rows of seats and a crackling loudspeaker, where the referee counted time on a wristwatch. All three sites returned the same answer: no profile. No record. No date of birth. No weight class. No trace that this person had ever existed inside any system at all. What kept me sitting there longer than necessary was not the emptiness. It was the small line at the bottom of the page: unclassified. A person had stepped onto a mat, taken punches, won, lost, and the system still did not know which box to place him in. The training ground is empty of spectators, but the pulse of the ball still echoes. In combat sports, the gap in the data echoes differently — it sounds like a door closing in front of someone trying to enter the market. And I realised I was looking at one of the most serious problems in professional sport, a problem rarely discussed because it has no images, no highlights, no scoreline to argue about. WHAT THE MARKET IS ACTUALLY BUYING The current transfer window does not run on phone calls. It runs on data files. A club in Asia opens the profile of a midfielder and sees thirty recent matches, minutes played, touches, distance covered, top sprint speed. A boxing promoter opens the profile of a fighter and sees win rate, knockout rate, average rounds, height, reach, age. A data brokerage opens the same profile a third time and sees something entirely different: the liquidity of the betting market around that name. Three users, three needs, and all three believe they are reading the same truth. That is where a great many mistakes in this industry begin. During a transfer window, money flows along probability. A club does not buy a player; it buys a distribution of possible outcomes. A promoter does not buy a fighter; it buys a forecastable cash flow. Demand for profiles therefore spikes precisely when time is shortest and pressure is greatest. And it is precisely then that people check the provenance of data least. Contract structure is still where the real story sits. Release clauses, payment schedules, image-rights splits, wage-bill architecture — those determine whether a player can leave, not the three-minute video circulating online. But the three-minute video is what gets shared. The distance between those two things is where the observer's craft still has room to live. I have watched many transfer windows from the observation seat. The pattern repeats almost identically: the rumour travels first, the contract arrives later, and by the time the contract arrives the old rumour has been remembered as established fact. Fans do not remember who was right. They remember who was loud. Against that backdrop, a fighter who does not exist in three major databases is not a minor detail. It is a readable signal, if the reader knows what he is reading. The problem is that most of this industry does not. THREE SCORING LOGICS COLLAPSED INTO ONE LABEL This is the point I consider most important in the whole story, and the most overlooked. Martial arts is not one sport. It is at least three scoring systems that differ in kind, and they are merged under a single label on data platforms. The first system is professional combat: boxing, kickboxing, Muay Thai, grappling, mixed martial arts. Here, outcome is the only currency. You win, you lose, or you draw. Finish rate, knockdowns, control time on the ground — these are meaningful measures, because the purpose of the contest is to finish the opponent or to score by round. The second system is sanda, the hybrid combat discipline of Chinese martial arts, permitting punches, kicks and throws but no ground fighting. It sits between traditional wushu and professional kickboxing, with its own ruleset, its own officials, its own scoring. The third system is taolu, the choreographed performance routine. Here the concept of finishing a bout does not exist. There is no knockout. There is no opponent to be dropped. Points are awarded for movement difficulty and performance quality. A taolu athlete can be world champion without ever having struck another human being in competition. These three systems share a word in Vietnamese. They share a word in English. They share a label in the databases. That is where the trouble begins. If you take a mixed martial artist's finish rate and apply it to a taolu athlete, you do not get a weak result. You get a meaningless result presented as a finding. If you use knockdown counts to evaluate a forms practitioner, you are measuring something that has never existed in their discipline. If you use ground control time to rank a sanda fighter, you are discarding the single most characteristic element of sanda: the throws. Let me state this plainly: the biggest risk in sports data is not missing data. The biggest risk is applying the wrong lens with complete confidence. An empty cell is known to be empty. A wrong cell is mistaken for truth. THE UNCLASSIFIED LABEL AND THE TRAP OF TAXONOMY Back to that line at the bottom of the page: unclassified. In modern data architecture, the classification label determines everything downstream. It decides which metrics get computed, which get discarded, which model gets run. A good system separates professional combat from sanda, and sanda from taolu, before doing anything else. A poor system calls all of it martial arts and then runs the model with confidence. An undefined label is the signature of a pipeline broken in the middle. The input is not entirely empty — it holds enough to assign a general tag. But not enough to assign a specific one. The system falls into what I call the empty-label state: it knows something exists, but not what. For a fighter, this state has very concrete consequences. He is not placed in any ranking. He does not appear on any scouting list. He is not invited to events. He becomes invisible not because he lost, but because the system has no drawer to file him in. In the transfer market, that invisibility converts into money very quickly. A fighter with no profile has no valuation. A fighter with no valuation has no agent. A fighter with no agent has no contract. The chain closes in silence, and nobody writes about it because there is nothing to write. No scoreline, no drama, no moment worth clipping. I once assumed this was a technical problem belonging to Western platforms. Then I realised it belongs to an entire region. Leagues across Southeast Asia, national federations, small provincial promotions — most of them do not transmit data outward. No recording software, no data-entry staff, no data-sharing agreements. Results exist on paper, in a notebook, in a referee's memory. When that referee retires, those results disappear. AN EMPTY PROFILE IS NEITHER A VERDICT NOR A CLEARANCE There is a second misreading I need to rule out immediately. When a fighter has no profile, the market's default reaction is to treat him as unverified, therefore high-risk, therefore not worth signing. That reaction sounds reasonable. It rests on the assumption that the existence of data equals verification, and the absence of data equals non-verification. That assumption is wrong at both ends. At the first end, plenty of fighters have complete profiles inflated by bouts against weak opposition. The industry has its own phrase for it. A run of ten wins over ten opponents who have never appeared in any ranking says little about real ability. The data is there. The meaning is not. At the second end, missing data says nothing about ability. It says something about readability. Sports data does not measure truth. It measures the readability of truth. An excellent fighter in a promotion that transmits no data looks identical to a fighter who has never competed, on a scout's screen. A null result must be read as a null result. It is not an affirmative safety check. It is also not a conviction. It is a silence that needs to be filled by direct observation, not by inference. In combat sports, that silence is far more dangerous than in football. A player without data still has hundreds of recorded matches, still has scouts attending in person. A fighter without data often has only a few dozen people who have ever seen him fight, and most of them are hometown spectators. The players run on the pitch, and I note the rhythm of each footstep — the memory of an observer who never sits still. That memory tells me most of the great scouting errors of the past decade did not come from misreading data. They came from having no data at all, and drawing a conclusion anyway. THE COST THAT NEVER GETS LOGGED I need to tell a different story to make this clear. On 27 June 2026, in Kazan, South Korea beat Germany 2-0. Kim Young-gwon scored in the 90th minute plus three. Son Heung-min scored in the 90th minute plus six. It remains one of the most shocking results in World Cup history. In the match data, one line records the saves made by goalkeeper Jo Hyeon-woo. That figure sits in every statistical sheet, is quoted in every commentary, and feeds into every subsequent scouting file. What sits in no statistical sheet is what happened fifteen minutes after the final whistle. I was standing in the mixed zone, and while every camera had turned toward the goalscorers, Jo Hyeon-woo sat alone in a corner of the dressing room, head in his hands, crying. He had been criticised all week beforehand, after the 0-1 defeat to Sweden on 18 June 2026. That criticism is not in his file. Neither are those tears. I wrote about that moment, and some colleagues called it sentimental. Looking back years later, I still think it was the most explanatory detail of the entire match. A goalkeeper doubted by an entire nation, playing the game of his life, then crying alone when nobody was watching. The data sheet recorded the saves. It did not record the cost of the saves. This is the foundational limit of every sports data system. It records events. It does not record the cost of events. And when a fighter steps onto a mat, cost is what determines how much longer he can keep doing this. WEIGHT CUTTING: THE THING DATA DESCRIBES WRONG, SYSTEMATICALLY Across the whole of combat sports, weight cutting is the most urgent medical risk category. It is also the category data misdescribes most often. The mechanism is simple. Fighters cut weight rapidly in the days before the weigh-in, mostly by dehydrating. The body is pushed into acute dehydration, then rehydrated across roughly twelve to thirty-six hours before competition. Documented complications in the sports-medicine literature include acute kidney injury, rhabdomyolysis, and collapse on the scale itself. On the record sheet, this fighter looks entirely clean. He made weight. There is no note. There is no flag. Data platforms have no field for it, because that field does not sell tickets. This is why I argue that any scouting system reading only results has a serious hole. It reads the outcome of fighting. It does not read the conditions of fighting. And in a sport where the body is the only tool of the trade, the conditions of practice matter as much as the results of practice. I have spoken with organisers at small promotions in Southeast Asia. Most of them have no weigh-in doctor, no medical committee, no mandatory suspension protocol after a knockout. A fighter can be knocked out in the first round and be back on a card three weeks later. Nobody logs it. Nobody tracks it. And when data does not log, nobody knows there is something to intervene on. THE SILENCE OF HEALTH DATA There is a misunderstanding I encounter constantly in this job, and I want to address it directly. When a fighter's profile records nothing about injuries, readers assume the fighter has no injuries. That is a formally reasonable inference and a practically false one. The absence of data is not evidence of the absence of a problem. In brain health, this problem is far more severe. Assessing long-term risk requires at least four groups of information: total head strikes absorbed, number of knockouts, concussion history and the return intervals between them, and sparring volume inside the gym. Of those four, only the second is even roughly traceable in public databases. The other three barely exist. Nobody counts head strikes in a training session. Nobody logs the interval between two concussions. And sparring volume lives inside a closed gym, no cameras, no spectators. When the stands are empty, I hear the players breathing more clearly than ever — a sound no television broadcast captures. But the breathing inside a morning sparring session is heard by even fewer people. That is where the human brain is quietly consumed, and it is where data never sets foot. The consequence is that fighter health records are fundamentally incomplete documents. Not false. Incomplete. And during a transfer window, an incomplete document is routinely read as a clean document, because nobody has time to ask what is not being recorded. WHO PAYS FOR LIVE DATA I need to give this topic its own space, because I consider it the darkest side effect of the digitisation of sport. Live data from competitions is sold to betting companies. This is no longer hypothetical. It is a standardised revenue line, written into contracts, negotiated like broadcasting rights. Every pass, every touch, every serve becomes a signal within milliseconds and is sold onward. The issue is not whether betting exists. The issue is who the data is designed for. When a data capture system is designed for viewers, it prioritises metrics that tell a story: goals, saves, dribbles completed. When that same system is designed for the betting market, it prioritises metrics that can be wagered on in real time: next corner, cards, throw-ins. Those two objectives are not the same. This feeds back into how matches are staged. Longer breaks to give bookmakers more time. Events split into more sessions to create more betting windows. And most importantly: the observer at the ground gradually becomes a data supplier rather than a storyteller. I once received an offer to send live notes from training sessions to a data collection outfit. They did not ask about how the players felt. They asked about timestamps and counts. I declined, and I know many others did not. When every touch of the ball has a price, the question of who gets to see it becomes an economic question rather than a journalistic one. CLOSED ECOSYSTEMS DO NOT PRODUCE STARS There is a model that repeats in many parts of the world, and I consider it a structural error. When a league is run on fixed invitations rather than open qualifiers, it can survive for years. It can have sponsors, television, prize money. It looks very much like a professional league. But it does not produce genuine stars, because stars are forged in the risk of elimination. This holds true in women's esports. In many markets, women's competitions are structured to guarantee slots to a fixed group of teams. The result is no open qualifiers, no pathway upward from below, no story of an unknown roster beating the champion to write itself into history. A closed ecosystem can sustain the people inside it. It cannot produce the next one. And in sport, long-term commercial value comes from the next one. I believe the same principle applies to combat sports. A promotion that only matches fighters on its own roster against each other creates a small, stable, predictable circle. It will not create the moment an entire generation remembers. And without that moment, data becomes the only thing left — which is to say, the least important thing. PROVING YOURSELF AFTER INJURY There is a demand I hear constantly in this industry: a fighter must prove himself in the first bout back from injury. I consider that demand cruel, and worse, counterproductive on technical grounds. After a long injury, the body has not returned. Reflexes have not returned. Tolerance for impact has not returned. Those things need real competition time to recover, and no gym drill simulates the pressure of a real bout. But external pressure forces the fighter to go all-out from the first fight back, to avoid being labelled finished. The result is a loop. Pressure brings the fighter back earlier than is safe. Returning early raises re-injury risk. Re-injury lengthens the layoff. A longer layoff raises the pressure on the next return. On the record sheet, that loop appears as an uneven run of wins and losses, and is read as decline. That reading ignores cause. A fighter back after eighteen months does not have the same body as a fighter who rested eighteen days. In football, graduated return-to-play protocols exist, with managed minutes and staged exposure. In combat sports, comparable protocols barely exist at the organisational level. There is a fight, or there is no fight. Which turns every comeback into a medical gamble presented as a professional one. FROM THE GYM TO THE DIGITAL WALLET To understand why data carries such power, you have to look at how money travels through this industry. At the source are gyms and youth development systems. This is where fighters are produced, and where the least money sits. A coach at a small gym teaches three classes a day and earns far less than a mid-tier fighter. In the middle are promotions and event organisers. This is where contractual power concentrates, along with the power to allocate slots and decide who appears on television. At the far end sit broadcasting, data, betting and consumption. This is where the largest money flows, and where fighters have the least control. Publicly reported revenue-share figures in the industry show a wide and stable gap. In major mixed martial arts promotions, the share of revenue going to fighters has typically sat below twenty per cent. At the top of boxing, elite fighters can take more than fifty per cent. Major team-sport leagues generally share around fifty per cent. Entry-level purses for a new fighter commonly sit between ten and twenty thousand US dollars per bout, while training, nutrition, medical and travel costs are usually borne by the fighter. These ratios are not technical details. They are the architecture of a profession — one in which most practitioners cannot make a living from the trade, while carrying the entire physical risk of it. During a transfer window, that architecture is obscured by noise. The big contracts make the front page. The small contracts go unmentioned. And most fighters are in the second group. WHAT OUTSIDERS GET WRONG Let me use this section for a common misunderstanding. Fans often believe data is objective and human judgement is subjective. From that they conclude that more data will produce less argument. Reality runs the other way. Data is the product of a chain of choices: which metric to measure, which definition to attach to it, when to record, who records, whether to publish. Every choice in that chain can be subjective. The output is a number that looks highly objective but is in fact a packaged judgement. Once a metric becomes standard, the judgement inside it disappears from view. People argue about whether the number is right, and never about whether the metric should exist at all. A second misunderstanding concerns combat metrics. Viewers often assess fighters by win rate. But win rate says nothing without knowing who the opponents were. A high win rate over weak opposition is worth less than a moderate win rate over strong opposition. A record sheet does not distinguish the two cases, unless the reader does that work himself. A third misunderstanding concerns time. Old data in combat sports is more dangerous than old data in many other disciplines, because a fighter's body is the only asset and that asset depreciates. A two-year-old profile describes a different person. During a transfer window, all three misunderstandings surface together and push the perceived value of data above its real value. WHAT I AM TRACKING NEXT Based on my experience covering matches, there are four signals I will be watching in the period ahead. First, the classification label. When a data platform begins separating professional combat, sanda and taolu into distinct groups, that is a sign data quality is improving at the root. Second, the publication date. A profile without an update date is a profile that cannot be verified. I will prioritise sources that show dates, that name the person entering the data, and that link back to the original bout. Third, the platform's funding source. A data site living on advertising has different incentives from one selling live data directly to the betting market. Both can be credible to some degree, but the reader should know whose product he is reading. Fourth, how many fighters have updated profiles this season. If that number rises, the industry is widening who can be seen. If it stays flat while the number of bouts grows, the gap between those who compete and those who are recorded is widening. I am the one who keeps the rhythm — not for the match, but for the side stories bleeding minute by minute. In this period, the largest side story lies with the people who have never been written into any page at all. When a system cannot read a human being, the question is usually reversed: what is wrong with that person. That framing has survived far too long. In the next transfer window, when you see a name with no profile, try asking a different question. Not what this person is missing, but what the system is missing in order to read him. The answer to the second question is usually more useful, and usually more worth writing.