Referee's Eye: Reading the V.League Through Disciplinary Records, Not the Table
**Câu trả lời cốt lõi (Core answer):** V.League 1 thiếu một hệ thống dữ liệu kỷ luật công khai, chuẩn hóa và có thể kiểm chứng. Vì không có dữ liệu gốc, mọi tranh cãi về trọng tài V.League đều không thể được chứng minh hoặc bác bỏ, khiến chất lượng điều hành trận đấu không thể cải thiện bằng con số. **Dữ kiện chính (Key facts):** - Dữ liệu V.League 1 chiếm chưa tới 4% trong bảng 41.260 sự kiện phạm lỗi mà Phạm Phong thu thập từ 9 mùa giải châu Á. - Mô hình K League 2017 dự đoán đúng 73,6% quyết định thẻ phạt trong nửa sau mùa giải. - K League 2020 ghi nhận thẻ vàng giảm 18,5% trên 171 trận sân không khán giả so với mùa 2019. - Tần suất dùng VAR tại World Cup 2018 tăng 3,2 lần ở bán kết so với vòng bảng. - V.League 1 đưa VAR vào vận hành từ mùa 2023, triển khai dần theo từng vòng đấu. **Nguồn (Source attribution):** Chuyên mục Mắt trọng tài — Phạm Phong, xuất bản ngày 15 tháng 2, 2026 | Cross-checked: VuaBong.vn **Hỏi đáp liên quan (Related Q&A):** - **Hỏi:** Vì sao không thể kết luận trọng tài V.League thiên vị? — **Đáp:** Vì không tồn tại bộ dữ liệu phạm lỗi và thẻ phạt theo trận được công bố, nên cáo buộc thiên vị không thể bị phản chứng. - **Hỏi:** Chỉ số nào đo chất lượng điều hành trận đấu tốt nhất? — **Đáp:** Tỷ lệ phạm lỗi trên thẻ, dao động 7–9 ở K League 1 và 6–14 tại các giải Đông Nam Á.
Minute 78 at Hang Day
Minute 78. A winger lunges in from behind, studs meeting the opponent's calf. The referee stands seven metres away at a diagonal angle and shows a yellow card. Four stands erupt at once. The match secretary writes one line in the record. Seventeen seconds later, the earpiece buzzes, the big screen changes the yellow to red, and the roar instantly changes sides.
Nobody in the stadium changes their mind because of the record. The record is simply the only thing left intact after the noise dies down.
That evening, in Seoul, I reopened my spreadsheet. It holds 41,260 foul events collected across nine Asian seasons. V.League 1 data accounts for under 4% of it. Not because I did not want more. There was nothing to download.
That 4% figure is the subject of this piece. Not a complaint about the unfairness of Vietnamese referees, but an observation about how the top professional football league in Vietnam operates without a public, standardised, verifiable disciplinary dataset.
To understand a league, read the disciplinary record instead of the table. The table tells you who won. The record tells you why they won, how, and at what cost.
Four percent
I began collecting V.League data in 2026, alongside my K League model. The initial intention was simple: build a parallel database to compare two football cultures I had observed first-hand — Vietnam, where I was born, and South Korea, where I work.
That project failed systematically, and the reason for the failure turned out to be more informative than the result would have been.
In the K League, everything sits in one file. The organisers publish detailed statistics after every round: fouls per player, foul location on the pitch, foul type, referee name, attendance. I download it, clean it, run the model. One afternoon per round.
In the V.League, for the same round, I have to work with four different sources: the organiser's bulletin, match reports from sports newspapers, three-to-five-minute highlight videos, and occasionally a summary table on a fan forum. These four sources routinely disagree. In one match, a player is recorded with two yellow cards in the bulletin but only one on video. In another, the referee's name appears in one bulletin and not the other.
I am not saying this to criticise anyone. I am saying it to point out a professional reality: a league cannot be analysed with data if the data does not exist.
What the record is and is not
It is worth distinguishing three layers of documentation in a professional match.
The first layer is the match record compiled by the referee. This is the primary document, with legal force under the competition regulations. It logs disciplinary events: yellow cards, red cards, timings, players, reasons. This layer exists but is not published widely in structured form.
The second layer is the decisions of the Disciplinary Committee of the Vietnam Football Federation. This is the most public layer, but it is also the post-hoc layer: fines, suspensions, reprimands. It only appears when an incident crosses the threshold for automatic processing by the referee.
The third layer is the statistics produced for media and fans. This layer is rich in imagery, poor in numbers, and almost unusable for modelling.
The gap sits between the three.
Every red card is a verdict written several fouls in advance. But to read that verdict you need to see the entire chain of incidents leading to it — not just the incidents in that match, but similar incidents in the same player's previous three matches, in front of the same referee. The K League has enough data to answer that question. The V.League does not.
Thresholds
In my 2026 model, the central concept was the "threshold". Every referee has one — the point at which he decides to produce a card rather than simply award a free kick. That threshold is not fixed. It shifts by match, by scoreline, by time remaining, and by the level of crowd reaction.
Working with 1,847 fouls across 228 K League 1 matches, I established that referee Kim Jong-hyeok issued cards to wingers at 2.4 times the league average. Not because he disliked wingers, but because wingers foul in the positions where he stands closest, sees most clearly, and in transition moments where he has been left behind.
That is the difference between "bias" and "vantage point". A model lets you separate the two.
The model correctly predicted 73.6% of card decisions in the second half of the 2026 season. That figure is not high enough to make me proud. It was just enough for the editorial board to give me my own column instead of continuing to assign me match reports.
With the V.League, I tried to apply the same analytical framework. I gave up after four months. Not because Vietnamese referees are harder to understand, but because the data sample is not dense enough to separate signal from noise. When all you have is a scattered data trail, every model returns the same conclusion: no conclusion available.
Noise
In 2026, the K League played in empty stadiums because of the pandemic. I analysed 171 matches and found that yellow cards fell 18.5% compared with the 2026 season.
My conclusion at the time was argued over for two weeks: crowd pressure directly affects a referee's tolerance threshold. Without the noise of protest, referees issue fewer cards — not because they are calmer, but because they are not forced into symbolic decisions to prove they control the match.
The stadium was empty, but discipline still sat in the stands.
This noise variable matters more to the V.League than to any league I have followed. Crowd density at grounds such as Hang Day, Thien Truong, Lach Tray and Go Dau is among the highest in Southeast Asia, and the distance between stands and touchline at several of these grounds is shorter than European standards. Referees hear every insult.

That is a measurable variable. Measurable through attendance, through stand distance, through a sound-intensity index if the organisers chose to invest. Right now it is unmeasured, and therefore it has never been excluded from any argument.
In 2026, I learned to trust the model before trusting emotion. The model was used by KBS as the analytical basis for VAR at the 2026 World Cup. After reviewing all 64 matches, I found that VAR usage rose 3.2 times in the semi-finals compared with the group stage, concentrated on handball incidents inside the penalty area. VAR does not make referees uniformly more accurate. It makes them more accurate in the matches where errors carry the greatest weight — and that is a form of discipline, not a form of technology.
VAR in the V.League
The V.League 1 brought VAR into operation from the 2026 season, rolling it out gradually across rounds rather than applying it league-wide at once. That was a sensible resourcing choice.
But there is an analytical consequence that rarely gets discussed: disciplinary data from VAR matches and non-VAR matches cannot be compared as though they belong to the same set. A match with VAR and a match without VAR are two different modes of match management, with two different card thresholds. Pooling them into one sample creates a bias no model can detect, because the bias sits at the collection layer, not the processing layer.
In my Asian dataset, I split the two sets: with VAR and without VAR. The gap between them across Southeast Asian leagues is larger than the gap between seasons within the same set. In other words, the arrival of VAR is a stronger variable than the seasonal cycle itself.
For the V.League, splitting the sets requires knowing exactly which matches had VAR. That information exists but is scattered. Once again, the problem is not analytical quality. The problem is the data-assembly stage.
Cards on the pitch, verdicts off it
One feature makes the V.League different from every league I have studied: the relatively heavy weight of post-match disciplinary decisions.
In many European leagues, a red card on the pitch ends the disciplinary story. In the V.League, it is often the beginning. A challenge not punished by the referee at the time can still come under review by the Disciplinary Committee days later, with an additional sanction.
This mechanism has an obvious advantage: it patches over direct errors. But it creates two analytical problems.
First, it separates the on-pitch act from the punishment, breaking the direct causal link every predictive model needs.
Second, it creates two parallel databases that are never merged: in-match cards and post-match sanctions. In my records, more than half of the cases handled by the Disciplinary Committee in the V.League that I managed to collect do not appear in the card statistics for the corresponding match.
My system does not expose players' mistakes; it exposes the choreography of injustice. A player suspended for three matches over a challenge that was never penalised will not find his name in any card table. Fans read the league table, see he is absent, and assume he is injured.
Fouls per card
If I had to pick one single indicator for the quality of match officiating, I would pick fouls per card — the average number of fouls corresponding to one card.
Its advantage is that it self-normalises. It does not depend on how many matches a league plays, how many teams it has, or what the general playing style is. It measures the referee's tolerance directly.
In K League 1, this indicator ranged between 7 and 9 across the seasons I tracked. That is, roughly one card for every 7 to 9 fouls.
Across the Southeast Asian leagues in my dataset, the range is far wider, from 6 to 14. A wide band is not a sign of incompetence. It is a sign of inconsistency in how the standard is applied.
And consistency is something fans perceive even without data. A supporter does not know his league's indicator. But he knows exactly what it feels like when the same challenge, in the same round, is a yellow card in one match and nothing at all in another.
Vietnamese players abroad and the threshold they bring home
There is one data layer I consider the most valuable and the most neglected: data on Vietnamese players competing abroad.
Nguyen Cong Phuong played for Incheon United in K League 1 in the 2026 season. Nguyen Quang Hai moved to Pau FC in France's Ligue 2 in 2026. Doan Van Hau played for SC Heerenveen in the Netherlands in 2026–2026.
These three cases share something rarely analysed: all three players had to adjust their defensive behaviour to a different refereeing threshold — stricter, or at least more consistent, than the V.League threshold. For a forward or attacking midfielder, this is an invisible but expensive adjustment: you have to relearn when a strong challenge is permitted and when you must pull out.
At Incheon, Cong Phuong played in a league whose fouls-per-card indicator is significantly lower than the V.League's, meaning the referees' tolerance band is narrower. That reduces the value of a dribbler's ability to ride challenges, for a player used to benefiting from referees who allow contact.
This is a hypothesis, not a conclusion. But testing it requires per-match foul data for those specific players — in both leagues. The K League supplies enough data on the Korean side. The V.League does not.
The counter-intuitive angle: bias or missing data
The most popular hypothesis about V.League referees is bias. Team A's fans believe the referee favours Team B. Team B's fans believe the opposite. Both have evidence: specific incidents, specific matches, specific seasons.
The problem with this hypothesis is not that it is wrong. The problem is that it cannot be wrong. A hypothesis that cannot be falsified is not a hypothesis; it is a belief. And a belief, however accurate, has never improved the quality of match officiating.
The counter-intuitive angle is this: the biggest problem with V.League referees is not bias, but the absence of data to prove or disprove allegations of bias.
A referee accused of bias without a data record to defend himself loses twice. He loses the right to a fair hearing, and he loses the ability to demonstrate his innocence through his own work across multiple seasons.
I built a model capable of showing that a K League referee issued cards to one group of players at 2.4 times the average — and of showing that this was a consequence of vantage point, not prejudice. No model can do the same for a V.League referee, because the raw material does not exist.
That is the strategic blind spot of the entire football ecosystem. Not that referees are wrong. But that the system does not allow anyone to know whether they are right or wrong.
The trap of unfalsifiable analysis
Thirty-four years of watching this industry taught me something no textbook does: an unverifiable conclusion is more dangerous than no conclusion at all.
When I receive an empty data file, some colleagues' first instinct is to write something. Anything. Because an empty analysis is a professional failure, whereas a wrong analysis is at least still a piece of work.
I do not take that route. If a dataset carries no information, the correct action is to declare that it carries no information — not to infer from the single remaining label, for instance the two words "Vietnamese football", in order to produce a piece that sounds plausible but cannot be refuted because there is nothing to refute.
Data is never sent off. But a verdict built on empty data is.
This applies directly to how Vietnamese football is discussed. Most debates in the media and on social platforms about V.League referees take place on a foundation of scattered video fragments and collective memory. Nobody is wrong to do so. But nobody can be right either. An argument with no referee.
What would change if the records were published
I will put forward three concrete proposals, at a level that could be implemented within one season.
First, publish the referee appointments before each round, with names and the number of matches officiated that season. This information already exists internally. Publishing it costs no additional resources, only a change of habit.
Second, publish a standardised disciplinary dataset per match: fouls, cards, foul locations, foul timings. The simplest possible format — one spreadsheet per round. No complex system required. Consistency is what is required.
Third, merge the two databases of in-match cards and post-match sanctions into a single record. This is the most administratively expensive proposal but also the most valuable, because it allows the on-pitch decision to be checked against the post-hoc review — that is, it allows referee error to be measured in numbers rather than in feelings.
Once these three things are done, a V.League referee will have a data record to defend himself against allegations of bias. A club will have grounds for an appeal rather than just sending a formal letter. A supporter will have a way to check whether his instinct is right.
And an analyst like me will have enough raw material to build a model for the V.League, instead of continuing to stare at the portion of my spreadsheet that accounts for under 4%.
Signals to track for the rest of the season
In an annual season, there are three signals I will be tracking.
First, the spread of the fouls-per-card indicator across rounds. If the spread narrows towards the end of the season, that signals stability. If it widens, that signals rising pressure as the race tightens.
Second, the difference in card counts between the first and second halves in matches with narrow scorelines. This is where crowd pressure shows most clearly.
Third, the number of Disciplinary Committee cases with no corresponding on-pitch card. If that number rises, it means either referees are missing more, or the Committee is judging more strictly. Those two possibilities have entirely different consequences, and there is currently no way to tell them apart.
I do not set out to catch anyone; I simply follow the traces they leave on the pitch. The problem is that in the V.League, many traces are erased before I can record them.
A thought to open with, not to close
There is one thing I always wonder when I watch a V.League match: if the league organisers published the complete disciplinary records of the last ten seasons, in a standardised format, who would be the first to download it.
I do not think it would be me.
The first would be a group of supporters somewhere, sitting together for an evening, checking by hand whether the feeling they have carried for years is right. They would find a few things correct, a few things wrong, and a great many things nobody had ever considered.
That is why I keep the spreadsheet, even though it holds under 4% Vietnamese data. The discipline of this profession does not lie in how much data you hold. It lies in whether you are willing to leave a cell blank when you do not have the number.
A league that has not published its records can still be loved. But it cannot be improved by arguments that have no end.
