V.League: The Gap Between Points and the Coefficient Chain the Standings Never Reveal

On the bus from Nha Trang back to Saigon in the summer of 2026, I wrote my first xG table by hand for 14 V.League clubs, back when nobody called it data yet. The standings at the time placed one team in the group fighting for a continental cup spot, even though their total xG was lower than the team sitting eleventh. In the other direction, the eleventh-placed team created more quality chances than the runners-up. By the end of the season the points still stood there like a verdict; the coefficient chain, nobody bothered to read.

I am not retelling that to show off a notebook. I am retelling it to explain why, every time the transfer window opens, I reopen the old database before I believe any offer. My method is not complicated: for every shot in a V.League season, I record the position, the angle, the type of shot, the situation that led to it, and the moment in the match. From that I calculate xG, the probability that a shot becomes a goal, then accumulate it into a chain by match, by month, by season. For pressing I use PPDA, the number of passes the opponent is allowed before each defensive action; the lower the figure, the more aggressively a team presses.

V.League imposes limits I always have to spell out. The league has only 14 teams and 26 rounds, a sample small enough that one heavy rain or one early red card can bend a whole month of data. Pitch quality, travel intensity between regions, and the gap in the number of foreign players in each squad are all variables my model does not fully measure. But precisely because the sample is small, I have to read the long chain instead of a single moment.

The first thing the xG table exposes in V.League is that the gap between process and result is far larger than in Europe's top leagues. In a 26-round competition, the champion is often only a few points ahead of fourth place, so one lucky shot or one controversial penalty can flip the title. I once tracked a team with a positive goal difference built on a run of three straight wins in which their total xG was actually lower than their opponents'. Those three matches were not long-term form; they were just noise.

The second thing, and the one I hold dearest: teams that are undervalued usually run a more stable operating model than the standings assign to them. The world looked at Croatia as an underdog; I looked at them as a coefficient chain nobody had dared to exploit, and I apply exactly that principle to V.League. There are teams with no stars and no budget, yet their PPDA stays low with persistence, meaning their pressing system is organised rather than driven by mood. That persistence, across 26 rounds, is worth more than one individual shining for five matches.

When the transfer window opens, I always split the data into two layers: the result layer and the process layer. A foreign striker who scores 15 goals but has an xG of only 9.5 has finished above the norm, and the norm will return. A striker who scores 7 but has an xG of 12 is being undervalued by the very number on the scoreboard. The V.League transfer market usually pays for the result layer and ignores the process layer, and that is where value is mispriced the most. The transfer market is a game for those who look far, not those who look much; value always arrives after patience.

I still keep the habit of scanning every young player through their xG-per-match chain rather than their goal tally. In 2026 my model surfaced a 20-year-old winger with 0.48 xG per match, higher than the average of foreign strikers in the league, even though he scored only 5 goals. Many people called me deluded by numbers. But the coefficient chain does not care about the criticism; it only cares about position and shot quality. A year later, that same player scored the decisive goal on the regional stage, and those who had laughed at me went quiet.

I remember one evening in the stands, in the rain, with the away team leading from the 20th minute. In the crowd's eyes it was a settled game. In my eyes, the leading team was letting the opponent create four clear chances inside the first half alone. Fans watch the play; I watch 22 numbers moving, and I wait patiently for them to tell a different story. The leading team won 1-0, but the match xG was 0.8 against 1.9. The result went into the history books, while the model recorded a warning for the return leg.

In defence I pay particular attention to set pieces. V.League is a competition where dead-ball situations account for a high share of goals, partly because the quality of zonal defending still varies widely. A team can defend well in open play yet concede steadily from corners. This is the kind of problem the standings do not expose but a set-piece breakdown does, and it is also the kind that can be fixed through coaching far faster than waiting for a striker to rediscover his scoring touch.

But I have to challenge myself right here. My model does not cry, does not celebrate, but after every match it owes me a lesson. Correlation is not causation. A team having a higher xG than its opponent does not automatically mean it will win, and a striker scoring more than his xG does not automatically mean he will decline. In V.League's small sample, a single injury to a key centre-back, one disciplinary ruling, or one VAR decision in a grey area of the law is enough to bend the entire chain.

VAR is the clearest example of the limits of a number. I once believed technology would make the data cleaner, but in reality it only moved the argument from the pitch to the review room. A goal disallowed for an ambiguous offside can turn a match with an xG of 2.1 into a defeat, and my model has no variable to measure the ambiguity of a drawn line. I do not trust the coach; I trust the model. But I listen to the coach to fix the model, because only he knows why a player runs into the right spot yet never receives the ball.

What I fear most is not a model that is wrong, but a model user who is too confident. I have lived with a hand-written xG table since early on, and precisely because of that I know how well a pretty number can lull a reader to sleep. I no longer write about instant data; I place it in the context of history, governance, and the variables a spreadsheet cannot touch.

If this transfer window has one signal worth tracking, it is the question every V.League club should ask itself before signing a contract: am I buying a coefficient chain, or am I buying a stretch of noise? The answer does not lie in the transfer fee, but in whether the club has enough patience to read the long chain. The first xG table I wrote by hand on that bus, back when nobody called it data, is still, to this day, the most honest mirror I have.