The Empty Data Sheet and the Habit of Filling It With Prejudice

In September 2026, a twelve-page PDF landed in my inbox, sent by a sports data analytics firm trying to break into the South Korean market. The file title read "Deep Report - WK League Club". Inside were pressing indices, expected goals, line-breaking passes, heat maps by pitch zone, and a ranking of each player's influence.

I scrolled to the appendix to look for the sample size. Four matches. The club had played twenty-one that season.

Four matches, and twelve pages of conclusions. Not a single line in that file noted that the reliability of the figures inside was too low to base any decision on. The format of the report itself - the tables, the terminology, the order of presentation - manufactured an authority the content did not possess.

Three years later, I received another document. This time it was a nine-dimension analytical report on an esports article, built on a professional framework: patch and meta, tournament format, rosters and players, regional landscape, club finance, rules compliance, risk profile, public narrative, and industry transmission. Nine dimensions. Dozens of tables. And in nearly every cell, the same answer: insufficient information to conclude.

That was the most honest document I have ever read in this industry.

To see why that matters, look at the data architecture of women's sport. The WK League, South Korea's top women's football division, was founded in 2026. The men's K League 1 was founded in 2026. That twenty-six-year gap is not only about time. It is a gap in infrastructure: cameras per match, motion-tracking systems, statistical staff, broadcast contracts. All of it started later, at smaller scale, and in several seasons did not start at all.

The consequence lies in how the analytics industry treats that void, and there are two ways, both wrong. The first is to ignore it: no data means no analysis, so women's football gets pushed out of every tactical segment. The second is more common and more dangerous: fill it in.

When the data is empty, the reflex of the media industry is not silence - the reflex is to fill it.

The material used to fill it is usually prejudice dressed as observation. "This team runs a lot but lacks organisation," based on three matches watched. "This player has pace but makes slow decisions," based on two remembered passages of play. In men's sport those sentences still exist, but thick data sets stop them before they reach the broadcast. In women's sport, they slide straight into the piece.

The most technical point here is sample size, and it is harsher than it looks. A WK League season gives each club only a little over twenty matches. In football, a player's conversion rate needs hundreds of shots before it settles; a goalkeeper's save rate needs thousands of faced attempts before skill separates from luck. With four matches of data, a striker can be labelled a clinical finisher or a wasteful one almost entirely by chance. The problem is not the small sample. The problem is a small sample presented in the language of a large one.

I tried to do it differently. Based on my experience tracking and hand-charting WK League matches and women's esports tournaments in Seoul, I found that notes written by hand - which player moves when the ball is elsewhere, who calls the rhythm, who stays calm in the eightieth minute - often yield more information than a metrics table computed from four matches. I learned to listen to what the pitch whispers when nobody is filming. But handwritten notes cannot be sold to a sponsor, and metrics tables can. That is the root of the problem.

One more thing needs saying plainly: data specialists are walking into the locker room, and their conclusions often drift away from the actual rhythm of the match. In sport, the most important match sometimes takes place behind the dressing-room door. I interviewed Kim Jung-mi, goalkeeper for Incheon Hyundai Steel Red Angels - seven-time consecutive WK League champions - during the pandemic shutdown. What she described exists in no metrics table: having to act as her teammates' psychologist, the fear of losing form while unable to play, the pressure of a team that won so much that losing became abnormal. No forecasting model captures that variable.

Back to that nine-dimension document. The point I want to stress is not its emptiness. The lesson is that the document called itself a process defect. It stated clearly: the input contained no analysable information; any conclusion about the subject would be fabrication, and therefore every conclusion was withheld. It even assigned itself the highest risk rating - not the subject's risk, but the risk of issuing confident judgements from an empty evidence base.

Put the two documents side by side: one is twelve pages of conclusions from four matches, the other is nine dimensions with dozens of tables and the answer "insufficient information" in nearly every cell. The second is more modest, and more useful.

The counterintuitive view here is this: the problem with women's sport is not a shortage of data. The problem is that this industry is far too lenient with data that has shape but no substance. Adding metrics to a thin sample does not reduce error; it multiplies the number of ways to be wrong. The industry rewards output volume, not the calibration of belief. A forty-table report looks more authoritative than a single line reading "we do not know" - even when that line is more correct.

The same holds in esports. Women's tournaments often do not publish match data; fans are left with video and commentary. What gets produced most in that vacuum is not analysis, but speculation decorated with jargon.

There is a paradox of commercial value and competitive value here. Whoever pays for a report is not paying for caution; they are paying for conclusions. But on a thin sample, one wrong conclusion costs far more than a refusal to conclude. The cost of caution is small and paid once. The cost of misplaced confidence is large, paid in credibility, in bad transfer decisions, in a player undervalued for an entire career because of four unlucky matches.

And here is the part I believe will change. The infrastructure is arriving. Women's football leagues in Europe already have motion-tracking; the WK League is being streamed more; women's esports tournaments are starting to attract sponsors and their own data sheets. The unannounced door often opens onto the biggest stadium - and in this case it is opening both literally and figuratively.

But infrastructure only solves half of it. The other half is habit. If this industry uses the extra data to produce more tables instead of saying "insufficient information" more precisely, reasoning quality will not rise - only volume will. The real measure of a mature analytics culture is not how many metrics it produces, but how often it dares to refuse a conclusion.