The Empty File: How Combat Sports Fills a Data Void With Guesswork

11:47 p.m. A 47-page document lands in my inbox, and I do what I have done for twenty years: I count information points. Not pages. Not tables. Only the cells that can carry a conclusion. There are none.

Eight analytical dimensions — technical matchup, fighter condition, organizational landscape, business model, ruleset, health risk, public narrative, industry transmission — all return the same line: insufficient information. Once for every cell. Forty-seven pages, zero usable value.

The person who sent that file did not ask me how to fix it. They asked: so what do you think will happen? That is the question an entire industry asks itself every day, and it is the question that has produced most of the errors I have spent a decade cleaning up. A data void always gets filled. The only question is what fills it.

Forty-seven fights, forty-seven pages

I started writing about sport in 2026 while living in Australia, later moved to Vietnam, and have sat in enough meeting rooms to know one rule: the market does not pay for emptiness. An assessment with a clear verdict always sells. An assessment that says there is not enough data gets filed as incompetence.

July 2026 was the first time I hit that rule head-on. While working as a commentator for a television station in Guangzhou, a club asked me to assess the injury profile of Brazilian striker Alan Carvalho before a long-term deal. I reviewed 47 of his matches across 18 months, cross-referenced with GPS training data. One detail would not leave me alone: on artificial turf, his sprint power dropped 15 percent. Fifteen percent is not an elegant number. It means that in the decisive moment of a move, he arrives roughly two steps later than the defender — just enough for a cross to be blocked, just enough for a hamstring to be stretched past its limit.

I advised the club not to sign him long term. Six weeks later, Alan Carvalho tore his hamstring against Shanghai SIPG.

Forty-seven real matches. Seven years later, in another city, I received a file that was also forty-seven pages long and contained no matches at all. The coincidence made me pause before replying.

The framework and its empty cells

The framework I use to decode a combat sports event has eight dimensions. It is not academic ceremony. It is a list of the places where truth tends to hide, built after being ambushed too many times by fighters who looked healthier than they were and matchups that looked more even than they were.

Based on my experience tracking fights, the technical dimension speaks fastest. A fighter landing 5.2 significant strikes per minute is placing roughly 15 clean shots per three-minute round — that number decides whether he can genuinely press forward or only looks good at range. The strikes-absorbed figure tells the opposite story: how much the body has already paid. With an empty source file, neither number exists. You cannot compare styles when you do not know who throws, who absorbs, and who scores takedowns per 15 minutes.

The condition and career-longevity dimension takes the most of my time. A fighter's age is not measured in birthdays but in rounds absorbed. A 28-year-old with 22 professional fights and two 10-percent body-mass swings per season is older than a 34-year-old with nine fights behind him. Making weight is the mandatory ritual of this sport and the point where the body pays the highest price.

In April 2026, a Portuguese fighter named João Carvalho, aged 28, died after a bout in Dublin. That death changed rules. The California State Athletic Commission subsequently rolled out a multi-step weight-management program, including out-of-competition weight monitoring and restrictions on IV rehydration. From 2026, a major Asian promotion enforced year-round weight rules with hydration testing, turning weight cutting from a private trick into a measurable variable.

Those changes exist because someone died. In an empty data file, they become a blank labelled insufficient information.

The organizational dimension works differently. Exclusive contracts, titles fragmented across sanctioning bodies, cross-promotion superfights blocked by clauses — these decide whether a fighter gets meaningful fights, not talent alone. With no organization named and no event named, there is nothing to analyze.

The business dimension is similar. Revenue in combat sports comes from four sources: gate and live attendance, broadcast rights, pay-per-view, and sponsorship. Public estimates over many years have placed the fighter share below 20 percent of total revenue — meaning that on a night generating 10 million units, the portion reaching the people actually being hit in the head may not exceed 2 million. That ratio determines how many fights a fighter must take per year to survive, and that density loops back and damages the health dimension. Fight density is the biggest single cause of injury, and no medical team can save a fighter on a two-fights-a-week schedule.

The rules dimension has a recent milestone worth remembering. The partnership between the world's largest MMA organization and the United States Anti-Doping Agency ended in 2026; from January 1, 2026, Drug Free Sport International took over the testing program. For a spectator this is a technical footnote. For a fighter considering a supplement, it is an entire career.

The health dimension bundles four inseparable risks: cumulative brain injury, weight-cut catastrophe, musculoskeletal injury, and financial security after retirement. The narrative dimension measures the gap between public expectation and what happens in the cage. The transmission dimension traces effects from gyms to broadcasters, betting data, equipment, and local policy.

When the source file is empty, all eight return the same result. A report that looks complete, professional, and has not a single bone in it.

Why an empty file still produces an analysis

The night in Kazan in July 2026 taught me this in a way I cannot forget. During the quarter-final between Brazil and Belgium, while nearly everyone in the stands and on the microphones waited for Neymar to explode after his foot injury, I presented data from 12 matches I had collected myself: his change-of-direction capacity in the second half was down 12 percent, and his left thigh responded 0.3 seconds slower. A 0.3-second delay is not an abstraction. It is the gap between touching the ball at the right tempo and letting it run past your foot.

Brazil lost 1-2. Neymar lost the ball repeatedly near the box. The Kazan night taught me: public opinion is noise, numbers are signal.

But Kazan taught me the opposite lesson too, and that one matters more today: my numbers existed only because I had spent three weeks reviewing footage. Without those 12 matches of data, I would have had no right to say anything at all. I would have been the seventeenth person guessing on air.

In combat sports the void is far larger than in football. A fighter competes three times a year. A fight lasts 15 minutes. There is no 38-round season to accumulate sample size. That is why this industry runs on story: a training clip, a weigh-in stare, a line at a press conference. Those can be signals, but they are not data.

The summer of 2026 is when I understood this completely. With the pandemic cancelling every commentary contract and arenas empty, I contacted 23 young fighters and players, collected sensor data from their home sessions by phone, and spent eight months building a load-recovery model. I tested it on my own body first, then on them. When competition resumed in June, the group I tracked recorded only four injuries across their first ten matches, roughly 30 percent below the average of the previous two seasons.

That model sat scattered across 12 spreadsheets and was never widely adopted, because long-term planning is not my strength. The 2026 spreadsheet taught me: the body never rests; it only needs an algorithm patient enough. It also taught me that a correct model without proper documentation becomes an empty file for whoever comes next.

Going against the current: two failure modes, not one

This industry has two failure modes, and they mirror each other.

The first is fabrication. Without data, people lend their reputation to a guess, and because guesses are written in a confident voice, readers have no way to distinguish them from a measured conclusion. This mode produces beautiful predictions, fluent transfer assessments, and a long list of mispriced fighters.

The second is paralysis. Declaring insufficient data for everything, including things with plenty of data, so as never to be held responsible for a wrong call. I have met colleagues who have written for twenty years without ever making a judgement that could be proven wrong. That is not caution. That is self-defence.

I try to hold the middle, and the way I hold it is an unwritten rule: I do not write a contrarian piece unless the data in my hands can reconstruct at least one testable causal chain. If I do not have that chain, I state clearly what I am missing and stay silent on the rest.

There is one uncomfortable detail outsiders rarely see. Given the same data file, two analysts can reach opposite conclusions and both can be right by their own data. The difference lies in background assumptions: one assumes a fighter recovers along a straight line, the other assumes the body heals in steps. No dataset announces that assumption. Only the writer does — or hides it.

And there are things data cannot see, however much I dislike admitting it every time.

What the data does not see

Sensors measure velocity, acceleration, heart rate, reaction time. Sensors do not measure a fighter sleeping four hours because his child is sick, or signing a contract his family opposes, or taking one last fight because of a mortgage payment. Psychology, culture, and personal circumstance are variables that live outside my spreadsheet, and they can outweigh every metric I collect.

A clean spreadsheet can make people forget there is a person behind it who can be broken. A body reader like me knows: every pain is an answer — but not every pain answers the question I was asking.

Signals to track next

Three signals I will be watching, none of them about predicting winners.

The first is empty data files being resubmitted in complete form. An injury file with a name, a date, a fight count, a weight-cut history, and actual minutes competed is the minimum starting point. When that signal appears, analysis that currently cannot be done becomes possible after a single correction.

The second is sources and publication dates being stated explicitly. A number without a source is not data; it is a rumour with a unit of measurement. When a sports platform starts demanding this in every article, the quality of combat sports debate will change within one season.

The third is entity lists being filled in: fighters, organizations, events, coaches, gyms. Without that list, every analysis is literature.

For the domestic combat sports market, where complete fight records are still scarce, the third signal matters most. Vietnamese MMA promotions are growing on far less data than fans demand. Whoever documents first speaks first — not because they predict better, but because they are the only ones who can say what I said to the person who sent that file: I do not have enough data, and I know exactly what is missing.

Injury data never lies; only impatient readers do. But patience does not mean waiting in silence. It means naming the empty cells and refusing to fill them with imagination — because in this sport, the person who pays for a wrongly filled cell is not the writer.