Reading Back 42 Crowdless Matches: Where Does Home Advantage Actually Live

That night I stayed at the Becamex Binh Duong office until nearly eleven. On screen was a spreadsheet of 42 V.League matches from the 2026 season, filtered clean of spectators — a set of conditions nobody expected would become a natural laboratory for football's oldest question: where does home advantage actually come from. The home win rate fell from 49% in the 2026 season to 38%. The coaching staff wanted to keep the same home-and-away game plan as before. I objected, and I put a comparison table on the table.

From a lecture-hall blog to the analysis room

In 2026 I was twenty, a student of International Communication in Binh Duong. I started a blog called "Numbers Don't Lie" and spent three months building a rough xG model in Excel to re-score the 26 rounds of the 2026 V.League. The result kept me up at night: Quang Nam FC won the title with an average possession of 48% — the lowest among the top five — yet a finishing conversion of 17.5%, among the highest in the league. I wrote a piece titled "The Champion Who Doesn't Need the Ball" and drew no small amount of mockery. Three months later Quang Nam were crowned, and the piece was shared more than two thousand times.

The first lesson did not lie in the result. It lay in the fact that I learned to open with xG, with finishing conversion and chance quality, instead of retelling a match by feel. Possession is an index of duration, not of quality — and duration has never scored a goal. A team holding 61% of the ball but generating only 0.89 xG in a big match is holding the ball in areas from which nobody scores.

From then on, every judgement I wrote had to be tied to a specific number, a specific sample size, and a specific context. When I finished writing, I checked the data again before publishing. That discipline took shape in the early years of the job, when I learned to watch a match before rushing to a conclusion about it.

The chain of evidence: when the ball no longer belongs to the host

In 2026, thanks to the blog's reputation, a sports outlet invited me to work as a data assistant for the World Cup in Russia. When Germany lost 0-1 to Mexico in the opener, I dug up their last four matches and calculated PPDA — the number of opponent passes allowed per defensive action. Mexico pressed ferociously with a PPDA of just 8.7. Germany needed an average of 11.3 passes to stop a single play.

I wrote "The Rusting Machine" before the final group match, pointing out that Germany's midfield produced a mere 0.89 xG despite 61% possession. Germany lost to South Korea and were eliminated. The piece reached 45,000 views. PPDA and ball-circulation speed are the earliest warning gauges of a system that is rusting — they speak before the scoreboard speaks.

I wrote about Germany's collapse before the tournament. It wasn't that I was clever, only that I didn't believe in the myth. When a world champion enters the next tournament, the public looks at the trophy while the data looks at passes per defensive action. Those two things rarely tell the same story.

By 2026 the pandemic closed the stadiums. Thanks to the World Cup series, Becamex Binh Duong hired me as a data assistant. And I discovered something that made the whole coaching staff sit down: home advantage almost entirely vanished without spectators. The empty stadiums of 2026 made me ask: does home advantage come from the pitch or from the crowd? The data has an answer.

Which gauge was warning

The 42-match spreadsheet gave me three layers of signal. The first was the home win rate: 49% down to 38%. The second was the expected-goal gap between host and visitor: the gap narrowed sharply with empty stands. The third was cards and contested decisions favouring the host: a clear decline.

Those three layers combine into a tidy hypothesis: most of home advantage lives not in the grass, not in travel distance, not in familiarity with the ground. It lives in noise. Noise acts on referees, on player psychology, and on the rhythm of the match. When the noise disappears, those three layers of effect disappear together.

I proposed switching to proactive defending away from home instead of keeping the old plan. The club won four of its next five matches. But I don't cite that number to praise myself. I cite it to make one point: a model has value only when it changes a decision on the bench.

Numbers don't lie. But reputation whispers into the ear of the one who doesn't read the table.

Correlation is not causation

Here I have to stop and steer my own argument back. The fact that home advantage fell when the stands emptied proves that the crowd is part of the cause. It does not prove the crowd is the only cause. A sample of 42 matches is small, and the 2026 season was further muddied by a compressed schedule, player fitness, and the uncertainty of a pandemic year. I always state the sample size, the assumptions, and the uncontrolled variables whenever I present a table.

There is another blind spot that pure data cannot see: dressing-room chemistry. Player-valuation models tend to overrate the potential of young players and underrate what happens inside the dressing room. A young, cheap signing with pretty numbers can break the balance of a squad that is functioning well. A spreadsheet cannot measure that, which is why I never let a single index judge a human being.

The same goes for the goalkeeper position. A goalkeeper's distribution is being sanctified. A keeper with lovely passing but declining basic reflexes can still be priced high in the transfer market, because distribution metrics are easier to count than hard saves. Being countable is not the same as being important.

And I don't forget a group the data overlooks: young players who develop early. They are pushed into the pace of adults before their bodies have matured. On the spreadsheet they look like talents ready to go. On the injury curve three years later, the story looks entirely different.

I hate uncertainty. But 2026 taught me that an unforeseen variable can be stronger than any algorithm.

Signals for the next round

The transfer market is full of names being paid for the past. I make a living reading the future. But reading the future does not mean guessing. It means building the right index, placing it in the right context, and accepting that every model has a breaking point.

Numbers don't lie. But they only speak when we are willing to read them in their own context — the pitch, the weather, the pressure of the tournament, and the life cycle of form. Remove the context, and the prettiest number is just a pretty number.

For the season ahead, the signal I am tracking is not the home win rate. It is the ball-circulation speed in the midfield of the highly rated teams, and their passes per defensive action. When those two indices worsen while the scoreboard still looks good, the system is rusting from the inside.

I started a blog from a lecture hall, believing the data would speak for itself. Eleven years later, I teach it to speak in words. And the biggest lesson remains the one from when I was twenty: don't believe a myth just because it is told well. Open the table.

I don't predict. I read the data and accept the consequences.