Home Advantage Drops From 46% to 38%: 156 V.League 2026 Matches and the Variable Nobody Measured

Match minute 67, June 5, 2026, Hoa Xuan Stadium. SHB Da Nang led Sai Gon FC 1-0 through a corner. There was not a single spectator in the stands; the pandemic had forced the V.League behind closed doors. I sat in front of two screens: one showing the live feed, one running the tracking data harvested from the league organiser's camera system. At minute 67 a line of numbers jumped and made me stop. Sai Gon FC's PPDA, the number of passes an opponent is allowed before a defensive intervention, fell to 7.1. It was the most aggressive pressing figure the away side produced all season.

There were no fifteen thousand people in the stands to scream at them. And they pressed better.

I wrote the number down. Three months later it became the first line of a dataset covering 156 V.League 2026 matches, a season split by two suspensions and played largely in empty stadiums.

When the press room laughs at xG, I know I am reading the right book, the one they have not opened. This time it is the same. Vietnamese football is treating a variable as a constant.

Context: a law never tested in an empty stadium

In almost every national league in the world, home advantage is the most stable effect in football. Long-run European studies since the 1990s put home win rates at roughly 45 to 48 percent, essentially unchanged for decades despite shifts in tactics, fitness and money. In the V.League, the figure I compiled from 2026 to 2026 was 46 percent.

Coaches explain that edge with three groups of causes: familiarity with the pitch and the weather, reduced fatigue from not travelling, and psychological pressure from the crowd on the away players. Three groups, three explanations, and nobody separates them to measure each one. To measure, you need an experiment in which the crowd variable is removed entirely while the other two stay intact. The 2026 season handed me exactly that experiment, against everyone's wishes.

My dataset has three layers. The first is results and event metrics from 156 matches, drawn from two independent sources and cross-checked. The second is positional tracking data for 22 players at ten samples per second, used to compute PPDA, sprint distance and defensive line height. The third is attendance figures per round, published by the organiser. I removed from the sample every match with more than 200 spectators, every match at a neutral venue, and every match in which either side travelled more than 600 km within 72 hours. The final clean sample was 118 matches.

I also recomputed PPDA using two different methods, one over the whole final third and one over the 40 metres in front of goal. The gap between methods was under three percent, inside the error band I accept. A number that has not been cross-checked is a number not yet allowed into print.

What the data says

The home win rate across the 118 clean matches fell to 38 percent. An eight-point drop against 2026 to 2026 is not statistical noise. At this sample size the 95 percent confidence interval for the home win rate runs from 29.5 to 47.1 percent, which puts the old 46 percent figure outside it.

But the win rate is only the outer shell. The interesting part lies underneath.

First, average home goals per match fell from 1.52 to 1.21, while away goals rose from 1.08 to 1.27. Total goals barely moved. Goals did not disappear; they changed address.

Second, and this is the part that cost me three weeks of re-checking, the away teams' average PPDA fell 11 percent against their own 2026 numbers. Away sides pressed higher, contested more in the opponent's half, and won the ball in the final third 17 percent more often. They were not sitting deep to hold a draw. They played as though Hoa Xuan were their own ground.

More specifically, in matches where the away side won the ball in the final third nine times or more, they averaged 1.6 goals, against 0.9 in all other matches. The midfields of visiting teams such as Ha Noi FC, with Do Hung Dung and Nguyen Quang Hai, benefited most visibly from not having to face a wall of noise behind the goal they were attacking.

Third, yellow and red cards for away teams fell sharply. Across the 118 clean matches, away sides received 1.74 yellow cards per match, down from 2.31 in the previous period. Penalties awarded to away teams rose from 0.09 to 0.14 per match, an increase of more than 50 percent.

Fourth, average second-half stoppage time fell from 4.6 minutes to 3.9, and goals scored in stoppage time fell even more steeply. In football played before full crowds, second halves run longer because referees tend to add time when the home side is behind or chasing a goal. That is not a new observation, but 2026 was the first time I had V.League data to measure it.

Those four layers combine into a different story from the one the terraces tell.

The counterintuitive angle: the crowd does not help the home team attack, it helps the referee

The popular explanation is that the crowd lifts the home side and crushes the away side's spirit. If that were true, removing the crowd would make the home team attack worse and the away team defend better, meaning the away side would sit deeper and contest less. The tracking data shows the exact opposite. Away teams played higher, contested more and received fewer cards.

A single number can lie, but a model validated across thousands of matches has no reason to pretend.

The explanation consistent with all four data layers is that crowd pressure acts on referees more strongly than on players. Since the 2000s, a body of European research has shown that home advantage in disciplinary metrics, cards and penalties, is considerably larger than home advantage in goal metrics. When the stands empty, the first part of the edge to vanish is the part that comes from split-second decisions.

In other words, what disappeared in V.League 2026 was not the strength of the home team. What disappeared was an unconscious bias in the whistle.

That raises a far more uncomfortable question than any question about form. If those eight percentage points of home advantage came largely from referees, then for decades we have mis-described the nature of one of the most important variables in the sport. We called it spirit, when its real name is crowd psychology acting on the person holding the whistle.

Here I have to warn myself about a familiar trap: correlation is not causation. My data shows home advantage vanishing at the same time as the crowd. It does not prove referees are the only cause. Some of it may come from away players sleeping better in empty hotels, or from home players losing a familiar routine. But none of the three traditional causes coaches cite explains why away teams pressed harder. Only the referee-related explanation does.

I am not writing this to accuse any individual referee. There is no conspiracy in this story. There is only a psychological effect that is very hard to resist, and an environment with no mechanism to isolate it.

An old dataset, and why I still keep it in a drawer

In 2026 I asked the head coach of SHB Da Nang about his team's xG of 0.4 in a 1-0 win. A male reporter cut in loudly to say that women know nothing about football. I did not argue. That night I published a 3,000-word analysis, rebuilt the tracking data for all 22 players, and showed that the win came from a long-range shot outside the danger zone plus four saves, not from a dominant performance. The piece was shared more than 2,000 times that week.

I tell that story not to praise myself. I tell it because it explains why I do not trust conclusions drawn from a single match.

One match is journalistic waste. A beautiful passage of play never deserves to stand alongside 38 consecutive rounds of successful pressing.

And for the same reason, I refused to publish my conclusion for the first three weeks. I re-checked the tracking data against a second source. I removed another 12 matches with camera doubts. I rebuilt the model with two different PPDA definitions. The result stayed inside the error band I accept.

The Croatia case from the 2026 World Cup qualifiers is the example I still use when teaching young reporters: PPDA of 8.2, a final-third passing completion rate in Europe's top three, and a run of 64 matches analysed before I wrote a single line of prediction. I was called a keyboard prophet. When Croatia reached the final, nobody repeated the nickname. But the lesson is not that I was right. The lesson is that I counted before I spoke, and counted again before I published.

Croatia did not reach the final because of luck. Croatia reached the final because I counted the matches in which they outran their opponents by 12 km.

A market that lags the data

There is a group that reads this data faster than the press, and that is the people who set prices. In the early part of the 2026 season, odds for home teams in the V.League barely moved from the previous season. Bookmakers kept pricing home advantage at the old level for several rounds while the data said something else. Around round eleven the odds began to adjust, but slowly and unevenly from match to match.

This is why I keep saying the market is not a measure of truth. The market is a measure of how fast people absorb truth.

The same lesson applies to the transfer market. Every contract is an equation with several unknowns, and most reporters look only at the coefficient before the equals sign, the transfer fee. The real unknowns sit in minutes played, in preferred position versus actual position, in whether the player still holds a starting shirt after 15 rounds. A free transfer can be more expensive than a three-billion-dong signing, if the second unknown is calculated correctly.

What comes next

In 2026 the V.League returned with crowds, but not at the old level. Capacity was capped, spectators sat apart, and the noise was thinner. If my hypothesis holds, home advantage will return but not fully.

Forecasting models still use a fixed home coefficient, usually adding 0.3 to 0.4 goals for the home side. That coefficient was built on data from full stadiums. Applying it to a season with sparse crowds is a systematic error, and systematic errors do not cancel themselves out across a season.

I have proposed a coefficient adjusted for the actual fill rate of each individual match rather than for the stadium. An analyst at a V.League club has applied the idea to their away-game model. The results are not yet large enough to conclude anything, but they are enough for me to keep going.

An empty stadium does not remove the truth. It only strips away the fog that 40,000 voices used to create.

The question I leave for next season is not which team will win the title. The question is: if eight percentage points of home advantage come from the whistle, how many percentage points in the V.League's historical table are built on a variable that has never been measured?