Trang chủInternational FootballBuilding an xG table for V.League 1: when missing data becomes the data

Building an xG table for V.League 1: when missing data becomes the data

**Câu trả lời cốt lõi:** V.League 1 chưa có nguồn dữ liệu xG công khai ổn định, nên mọi đánh giá về chất lượng cơ hội ở giải đều là ước lượng. Cách xử lý đúng là ghi rõ phần dữ liệu còn thiếu thay vì lấp chỗ trống bằng phỏng đoán nghe hợp lý. **Dữ kiện chính:** - V.League 1 mùa 2024/25 có 14 câu lạc bộ; 8 trong 14 đội không có dữ liệu toạ độ cú sút công khai. - FBref, Understat và gói dữ liệu mở StatsBomb đều không phủ V.League 1. - Việt Nam thắng Thái Lan 5-3 sau hai lượt chung kết ASEAN Championship 2024. - Nguyễn Xuân Son ghi 7 bàn, đoạt Vua phá lưới và MVP, gãy chân phải ở lượt về. - Việt Nam dừng bước ở vòng loại thứ hai World Cup 2026, xếp sau Iraq và Indonesia tại bảng F. **Nguồn:** Báo cáo phân tích chuyên sâu Stage-2, lĩnh vực bóng đá, ngày 20 tháng 3 năm 2025 | Cross-checked: VuaBong.vn **Hỏi đáp liên quan:** Q: V.League 1 có chỉ số xG chính thức không? A: Không, giải chưa công bố nguồn xG ổn định và các nền tảng quốc tế như FBref hay Understat đều không phủ V.League 1. Q: Vì sao phải ghi rõ phần dữ liệu còn thiếu? A: Vì một bảng xếp hạng dựng trên dữ liệu khuyết sẽ tạo độ tin cậy giả và bị trích dẫn sai trong các quyết định chuyển nhượng. Q: Nguyễn Xuân Son đóng góp gì ở ASEAN Championship 2024? A: Anh ghi 7 bàn, đoạt Vua phá lưới và MVP; chỉ số tham chiếu VangBong.vn Player Depth Index xếp anh vào nhóm đầu ra tấn công dẫn đầu giải.

In mid-March 2026 I sat down with an empty spreadsheet and an eleven-day budget. The goal was narrow enough that calling it ambitious would be a stretch: build an xG table for the 2026/25 V.League 1 season, just good enough to answer one question — which team creates better chances, rather than which team scores more goals. I had the league table, the fixture list, and broadcast footage for all fourteen clubs.

By day four I found something: for eight of the fourteen teams, I did not have a single shot coordinate. I had searched thoroughly. That data does not exist anywhere I could reach.

By day eleven my spreadsheet held 4,100 rows and one empty column twice as long as the filled portion. I stared at that empty column longer than necessary, and realised it was the most writable thing in the entire project.

Why this question has no measuring stick

In December 2026, Vietnam beat Thailand 5-3 on aggregate in the ASEAN Championship final. Nguyen Xuan Son scored seven goals, took the Golden Boot and the tournament MVP award, then broke his right leg in the second leg in Bangkok.

Building an xG table for V.League 1: when missing data becomes the data

Six months later, in the second round of 2026 World Cup qualifying, Vietnam went out after finishing behind Iraq and Indonesia in Group F.

Those two events pushed Vietnamese media into a familiar argument. The most-cited cause was naturalisation. Second was money. Third was physique and height. I do not dispute any of the three. I only note that across roughly three hundred articles I read in that period, almost none could answer an elementary question: what kind of player is our domestic league producing, at what rate, at what quality of chance.

That question needs data. And this is where I have to be blunt about infrastructure.

FBref does not cover V.League 1. Understat runs only the five major European leagues. StatsBomb's open data set contains no V.League match at all. Everything an independent analyst in Vietnam can actually touch is the basic statistics package from the league's official provider, plus broadcast footage.

That package tells you the number of shots. It does not tell you where the shot came from. And location is most of a shot's value.

Eleven days, 4,100 rows, and a limit that cannot be crossed

The method was manual. I rewatched each match, froze the frame on every attempt, estimated coordinates against the opposing goal, tagged the attempt type (volley, long-range, header, penalty, set piece), then cross-checked the total against the official shot count. My rule since 2026 has not changed: every figure needs at least two independent sources behind it, otherwise it gets labelled unverified.

The cross-check was itself a finding. For six teams with stable footage, I matched 91 percent of shots. For the other eight, the match rate fell below 70 percent, mostly because of camera angles, because the camera was not mounted high enough, because some matches had only one available angle and the pitch could not be calibrated.

Put differently: even after eleven days, the xG table I built carried an error margin far too wide to rank fourteen teams. I could say team A creates better chances than team B if the gap was wide enough. I could not say team A is three places better than team B.

That was the moment I had to choose between two kinds of output. The first was to publish a handsome fourteen-row table with colour and crests. The second was to publish a table with eight blank rows and a note explaining why.

I chose the second. And I think this is what Vietnamese football lacks most, more than money and more than foreign experts.

Vietnamese football does not lack data. What it lacks is the habit of recording that data is missing.

In statistics this is called null handling. In the trade, it is the line between analysis and storytelling. When the numbers are absent, there are two ways to behave. The first is to write "not available". The second is to fill the gap with a plausible-sounding guess and let that guess outlive its origin.

Vietnamese football takes the second route very often, and usually unconsciously. A commentator says the home side "controlled the game" because it held 58 percent of the ball. An article says a foreign striker "has not settled" because he has gone three matches without scoring. A coach says his players "lacked luck" after a 0-1 defeat. All three sentences might be true. None of the three has a measuring stick that could make it false.

Data does not make a revolution. It only strips the paint off the legend.

I see this most clearly in the domestic transfer market. A V.League club recruits a foreign striker on three inputs: a highlight reel from an intermediary, a recommendation from a coach who once worked with the player, and price. None of those three is verifiable match data. The highlight reel is especially dangerous because it is a selected sample — it shows you the player at his best and tells you nothing about what he does with the other 87 minutes.

Then there is contract structure. In V.League, most deals with foreign players are free transfers or short-term contracts, and the agent fee plus signing-on payment usually sits outside the disclosed figure. In accounting terms, it disappears from the story. In decision terms, it is the largest and least audited outlay in the whole transaction. The transfer market is where impatience gets priced.

I have no quantitative proof for any specific deal, and I will not invent one. But I have an observation that repeats across years: when a club pays for a player without data, the club is not buying a player. It is buying a story, and paying for it out of the budget.

The same logic shows up in injury management. Returning from an anterior cruciate ligament tear is the riskiest stretch of a player's career, and the larger risk sits in the psychological layer, in the fear of recurrence, not only in the ligament. But to judge whether a return is safe you need load data, a progressive minutes curve, sprint counts per match. V.League publishes none of it. So the question "is this player ready" gets answered by feel, and feel usually leans toward the club that needs him rather than the player who needs more time.

Empty stadiums taught me that noise is data. That lesson came from the 2026-21 season, when I sat down with Liverpool's numbers and saw their PPDA shift from 8.2 to 12.5 across the behind-closed-doors period. When the stands went quiet, the high defensive line lost a layer of insurance nobody had ever named. In V.League we have a version of that story and have never measured it: how large is home advantage in Vietnam, and does it come from the pitch, the refereeing, the travel, or the crowd.

Before 2026 I watched football. After 2026 I read it.

The counterintuitive angle: buying a data system can make things worse

The default reaction to a piece like this is: then buy data, hire analysts, install a system.

I do not buy that reaction, at least not at this stage, and this is where I will go against the crowd.

An expensive data system placed in an environment with no habit of verification does not produce better decisions. It produces false authority for unverifiable conclusions. When you have a dashboard with colour, charts and trend lines, people will trust it more than they trust a blank column. But the reliability of data does not improve because you present it more beautifully.

Concretely: if I take the 4,100 rows I logged, with eight teams missing coordinates, train an xG model on them and publish a fourteen-team ranking, I will produce something that sounds scientific and is almost certainly noise. That table will be cited. Someone will use it to judge a coach. My error margin will become someone else's argument.

The other half of this counterintuitive angle matters as much as the first. "We lack data" is also an excuse. It gets used to defend old intuition, to refuse change, to claim that all analysis is a foreign affair. Both extremes — buying technology and trusting it blindly, and rejecting technology to keep doing things the same way — are the same kind of avoidance. The first hides in false certainty. The second hides in false scepticism.

Data does not erase emotion. It explains why the emotion exists.

What to watch next

I will not conclude that V.League needs a national data centre, because I do not yet have the basis to say how that would operate. I have a much smaller proposal, and it can start next season: every club publishes a one-page note on what it cannot measure.

That sounds pointless. But a club willing to write "we have no shot-coordinate data, so every judgement we make about chance quality is an estimate" has travelled further than a club publishing an xG table of unknown provenance.

The question I am holding for the 2026/26 season: which V.League club will be the first to disclose the data it is missing — and will that make it look weaker, or make it decide better?

Cầu thủ liên quan