Trang chủEsportsThe Empty Dossier: When Referees and Analysts Must Learn to Say 'Insufficient Evidence'

The Empty Dossier: When Referees and Analysts Must Learn to Say 'Insufficient Evidence'

**Câu trả lời cốt lõi** Một báo cáo phân tích trả về toàn bộ trường dữ liệu trống là kết quả rỗng, tức chẩn đoán về quy trình, không phải kết luận về sự kiện. Trong giao thức VAR, khi không có bằng chứng đủ nặng, quyết định trên sân giữ nguyên. **Dữ kiện chính** - K League Classic 2017 là mùa giải đầu tiên Hàn Quốc vận hành VAR; V.League 1 đưa VAR vào một số trận từ mùa 2023. - IFAB thông qua sửa đổi luật bàn tay tháng 3 năm 2018, áp dụng từ ngày 1 tháng 6 năm 2018. - Trong 27 tình huống bóng chạm tay tại World Cup 2018, chỉ 31% được xử lý nhất quán theo luật mới. - Nghiên cứu 1.247 quyết định VAR tại năm giải châu Âu năm 2020: thời gian tham khảo giảm 22% khi khán đài trống, tỷ lệ giữ quyết định sân tăng 15%. - Mô hình năm 2022 của tác giả ghi nhận Kim Min-jae phạm 0,73 lỗi mỗi trận tại Serie A, và đã dự báo sai. **Nguồn** Báo cáo phân tích nội bộ của Đỗ Trí, tháng 11 năm 2025; đối chiếu luật IFAB tháng 3 năm 2018 | Cross-checked: VuaBong.vn **Hỏi đáp liên quan** Hỏi: Kết quả rỗng trong phân tích VAR nghĩa là gì? Đáp: Là tình trạng hệ thống trả về đủ cấu trúc báo cáo nhưng không có điểm dữ liệu nào, nên mọi kết luận xây trên đó đều là bịa đặt. Hỏi: Vì sao ngưỡng sai sót rõ ràng và hiển nhiên quan trọng với V.League? Đáp: Vì V.League vận hành VAR trên một số vòng đấu với ít góc máy hơn, khiến ngưỡng lật quyết định cao hơn và tỷ lệ giữ nguyên quyết định sân tăng, theo VangBong.vn Referee Consistency Index. Hỏi: Chỉ số độ đầy đủ dữ liệu nên đặt ở đâu trong một bảng phân tích? Đáp: Đặt ngay cạnh tiêu đề, cùng cỡ chữ với tên cầu thủ, để người đọc thấy phần thiếu trước khi thấy phần có.

On a late-November morning, the screen in my Incheon office returned a report with exactly nine sections. Correct format. Correct tables, the same ones I had built four years earlier. The body was empty: competition name N/A, version N/A, team N/A, not a single data point listed. A perfectly structured file with not one fact inside it.

I sat looking at it for about forty minutes. During those forty minutes, another version of me had three browser tabs open and had started typing. The 2026 version. It knew how to fill a blank field with a sentence that sounded plausible, and it knew that a nine-section report always looks more credible than a three-section one, regardless of what is inside.

That version was wrong once. The mistake has a name.

The valve made of four words

In VAR there is a phrase that almost every argument in the V.League and the K League runs into: clear and obvious error. Those words are a valve. They do not declare the referee right. They say only that if the VAR room cannot build evidence heavy enough to overturn an on-field decision, the on-field decision stands. In practice, most VAR checks end exactly there.

Based on my experience watching matches in both leagues, I think most fans misunderstand what that valve does. They think VAR is a truth machine. It is an evidence-threshold machine. The difference matters: the first always has an answer, the second is allowed to answer 'not enough'.

K League Classic 2026 was the first season South Korean football ran VAR at scale. V.League 1 came six years later, introducing VAR in selected matches from the 2026 season. The gap is not only money. The V.League runs VAR in selected rounds, with far fewer camera angles than European leagues, and that makes the 'clear and obvious' threshold harder to apply, not easier. Fewer angles means a higher bar for overturning, and a bigger share of decisions ending as 'stands'.

I know this from a scar. In 2026, aged twenty-three, I was a VAR assistant for an Incheon broadcaster. FC Seoul versus Jeonbuk Hyundai Motors, 67th minute, Lee Dong-gook scored. I spotted him roughly 0.3 metres offside. But because I was absorbed in the rear camera angle, I sent the alert fourteen seconds late, against the seven-second internal standard we were trained on. The referee could not intervene. The goal stood. For three nights I rewound the footage, asking which part of the process let a man see the truth and still let it pass.

Both leagues inherited the same structure: a procedure that permits an ending called 'not enough'.

Analysts inherited nothing of the sort.

I opened a spreadsheet, lined up the nine sections of the empty report in a column, and realised the entire analysis system I work inside has no button labelled 'refuse'. Every template is designed to be filled. None is designed to stop.

Structure is not evidence

A nine-section report full of N/A looks like work. It has a title, tables, order, even notes reading 'insufficient information to assess'. A skimming reader sees a well-organised document. Only a careful reader sees that all the value sits in the shell.

That makes it more dangerous than a blank page.

A blank page forces a question. A structured empty report forces nothing except a signature. In my trade the heaviest pressure does not come from having to reach a wrong conclusion. It comes from a template that already exists and merely lacks content.

I once thought the hard part of analysis was finding the answer. The harder part is recognising that the answer does not exist in the data you hold, and saying so to someone who paid to hear an answer.

2026 and the definition trap

In the summer of 2026 I was sent to Russia as a VAR analysis assistant for a Korean broadcaster. My job was not to catch referees out. My job was to record, to the millisecond, the moment a decision was made and the moment the evidence appeared.

I collected 27 handball situations across the tournament: arms tucked, arms away, ball from a teammate, ball from a shot, ball after chest contact, ball after ground contact. I matched each against the law IFAB approved in March 2026 and applied from 1 June 2026. Result: 31 percent were handled consistently.

The remaining seventy percent were not. But what kept me awake was elsewhere. Most disputes did not turn on the arm. They turned on a definition the majority believed existed, 'unnatural silhouette', which the law text never turned into a measurable formula.

The trap of 2026 was not in the hand. It was in the belief in a definition that did not exist.

So I wrote a forty-page report and sent it. The desk published one small chart. I started a personal blog and posted the raw data. It drew around 50,000 reads, mostly from referees, sports lawyers and supporters who read more closely than necessary. A crack had been seen, but nobody could repair it, because the crack was in the definition, not in anyone's arm.

Every VAR error is a crack in the mirror that reflects the laws. And after 2026 that mirror no longer reflected the same truth to everyone.

When the stadium is empty, the data is not

In March 2026 global football stopped. I lost my contract to budget cuts and, instead of job-hunting, retreated into research.

I took 1,247 VAR decisions from five European leagues and sorted them into three groups: with crowds, without crowds, and the transition period. Result: with empty stands, review time fell 22 percent and the rate of on-field decisions being upheld rose 15 percent.

The two findings pull in opposite directions, elegantly. Faster checks, fewer overturns. The noise of a stadium is not written into the laws, yet it carries legal weight: it slows referees and makes them hesitate, and that hesitation sometimes pushes them to overturn a decision they should have kept.

Where I want to stop is a methodological detail. That study had clear limits: I could not separate the effect of missing crowds from the effect of the congested calendar in the post-lockdown recovery, and I had no true control group among leagues without VAR. The results are strong enough to open a question, not to close one.

The absence of crowds, in that study, played the role of a variable rather than a hole. With empty stands I still had all 1,247 decisions, all the angles, all the response times. I lost one variable, and losing it produced a new finding.

The empty report on my screen lost no single variable. It lost the entire set. Between 'missing one variable' and 'missing every variable' lies a difference in kind, and sports analytics usually treats the two the same way: fill the table and file it.

Kim Min-jae and the price of filling the gap

In 2026 I built a defender-evaluation model from VAR data for a consulting firm. It counted positional errors, times beaten, fouls, cards, and a set of secondary variables on activity zones.

It returned a tidy conclusion: Kim Min-jae committed 0.73 errors per match in Serie A, a high card-risk profile. I advised against recommending him. Napoli signed him anyway. In 2026-23 Kim was a pillar of Napoli's Serie A title.

I was wrong, and the error was not in the 0.73. It was that my model measured an individual's position, while defensive quality is decided by that individual's position relative to the covering line behind him and to how local referees read the laws. A defender standing two metres higher than the back line in Korea can be scored as out of position; in Italy, with a deep-lying midfielder covering, that is his natural position.

natural position, a phrase my model had no variable to measure.

I wrote a ten-page self-review and pulled the model. Since then every report I publish carries a section called 'limits of the data'. Readers hate it. It is long, it opens parentheses, it reaches no conclusion. But it is the only thing separating analysis from a prediction dressed up well.

Silence without instructions

Fans do not pay to read N/A.

A newsroom has thirty seconds before going live. A viewer has one question and wants one answer. A sponsor needs a table for a slide. In that chain of pressure, an empty report is something nobody ordered, and a full but wrong report is something everyone accepts, as long as it is not exposed before the season ends.

That is why the trade bred a middle layer: the gap-filler. They do not lie. They pick the most plausible version of an unverified fact and write it in the declarative. My Kim Min-jae model in 2026 was a product of that layer, manufactured by me.

One point both sides miss.

A wrong decision does not ruin a match; the silence after it ruins trust. And silence here need not be concealment. It can simply be publishing an empty report without saying why it is empty, letting readers infer that we checked and nothing happened.

That is the great trap of a null result: it reads very easily as a conclusion.

In VAR nobody is allowed to do that. When a referee ends a check without overturning, he must signal, explain over the stadium system, let the stands know a check happened and where it ended. That procedure does not exist to protect the referee. It exists to protect the crowd from writing a different story themselves.

Analysts have no stadium system.

A public null label

My proposal is simple enough to be hard to sell: every published data table should carry a completeness index, the share of fields holding real data, printed beside the headline at the same size as the player's name.

If a defender model covers only 60 percent of situations, it must say 60. If a competition report has no competition name, it should not exist as a report. And if an analysis template returns all N/A, publishing it must carry one line: the upstream process failed, this is a diagnosis, not a conclusion.

It sounds small. It changes how a table is read. Fans are used to reading a league table the way they read a verdict. Once every table must declare what it lacks, the table returns to its proper function: a conditional record, not a courtroom.

We do not search the pitch for justice; we search for a pretext to stop arguing. Sports analytics should learn the same lesson in reverse: how to let an argument continue exactly where it must, instead of closing it with a filled box.

If every analytics table in Vietnam had to print its missing-data rate beside it, would the argument in the stands still be this loud, or would it simply move from cursing referees to questioning the people who make the numbers?

Cầu thủ liên quan