When Vietnam's Volleyball Stat Sheets Contain Only the Score

I stayed behind with a sheet of squared paper after a domestic women's volleyball match. On the table were four things: the per-set score sheet, the registered line-ups, a ballpoint pen, and ninety minutes of footage. What I wanted to do was small: rebuild the first-pass efficiency of the two main attackers, count the rallies that fell out of system, and note which rotation let the opponent score in succession. Two hours later I had the score of every set. Nothing else. Nobody had counted before me, so I had to start from zero — and by the time I finished counting, the team had already moved on to the next round.

I do not trust the eye on the first viewing; I trust the third rewatch. But there is a limit the eye cannot cross: if nobody at the court records data on the spot, the analyst afterwards has only one option — to reconstruct the entire match by hand, rally after rally.

At world-level competitions, the FIVB operates a match information system that records data by rally and by athlete. A match there produces six core data groups: serve, first pass, set, attack, block, dig — plus errors. Every contact is assigned to a shirt number and an outcome, so that afterwards you can trace back: which rotation leaked, which attacker scored off a perfect ball, which attacker had to hit a bad one.

Volleyball rules allow a maximum of three contacts before the ball crosses the net. That limit is small, and precisely because it is small the causal chain inside each rally is very short: a first pass two metres off target, the setter loses the ability to run a fast ball, the attacker faces two blockers. Nine seconds later the crowd sees a failed rally. The spectator concludes the attacker is weak. The analyst with data concludes the first-pass system has collapsed.

On domestic courts, what survives a match is usually just the scoring report. That report is enough to confirm a result and to rank teams, but not enough to answer a single tactical question. Which rotation loses the most points? No figure. What percentage of perfect balls does the main attacker hit? No figure. Did this team win because its attack was good, or because the opponent self-destructed? No figure. A coach who wants to test his own hypothesis after the match has only one route left: reopen the video and count.

What the score alone hides

Start with rotation. A volleyball team has six rotations, and no two are alike. Some concentrate two strong attackers in the front row, some leave only one, and some leave the front-row attacking capacity close to zero. When a team concedes four points in a row inside a set, the first thing to check is which rotation it is standing in, not who just hit the ball out. Without recording rotation rally by rally, every post-match conclusion is a guess wearing a suit.

Next comes the first pass. In volleyball, the quality of the first contact decides the entire attacking menu behind it. If the ball is delivered to the right spot, the setter can run quick attacks, back-court balls, back-row options. If the ball drifts off target, the setter has one choice left: push it to the wing and let the attacker solve it alone. Those two situations produce two completely different attacking outcomes, yet on the scoring report they look identical — both are logged as one attack.

A stat sheet that contains only the score is never neutral. It assigns systemic failure to individuals, praises an attacker lucky enough to be served perfect balls, and stays silent about a block that has been stretched out of position for two sets.

In my recent work I have been forced to do something football abandoned long ago: build the data myself, from video. My channel is called Decoding the World Cup, but the counting method has no sporting borders. I pick one match, watch at 0.5 speed, and log every rally into a manual spreadsheet. A five-set volleyball match can generate more than two hundred rallies. Entering two hundred rows for one match is an evening's work, plus another evening to cross-check. Over a long season, that method does not scale. It is only enough to prove that data would tell a different story from the score — and that we are ignoring that story.

The blind spot is not in the software

The first reflex when people talk about missing data is to propose buying a system. That framing is wrong, and wrong in a dangerous way.

The bottleneck sits with the person at the table, not with the software. A match information system is worth exactly as much as the person entering data at courtside, and that person must be trained to grade a first pass as perfect, acceptable or poor consistently across an entire season. If today's recorder understands one thing and tomorrow's understands another, the dataset looks highly professional but cannot be compared across two rounds.

The second danger is importing definitions. The first-pass grading scale used at top-level competitions was built to serve setters of that level. There, a pass is considered perfect when it unlocks almost the whole attacking menu. Apply that scale unchanged to a domestic match, where the setter has fewer options and the ball travels at a different speed, and most genuinely good passes get marked as average. The numbers are technically correct but semantically wrong. We import the ruler and forget to import the table.

The third danger is harder to say out loud: data only has value when someone makes a decision from it. If a head coach already trusts his own feel, an extra data table just creates one more thing to ignore. A team can record all six data groups every match without changing a single rotation.

A rotation only becomes clear when you count the rallies that died in it, not when you look at the points the team scored there. Two teams can both lose a set 20-25 while committing two entirely different kinds of error, and the scoring report cannot tell them apart.

What can be done now

I am not proposing a large system. I am proposing four columns, one person, one tablet, and a grading rule written on paper before the first round of the season.

Column one: the rotation at the moment the rally begins. Column two: first-pass quality, in three grades. Column three: the outcome of the rally, including who the error belonged to. Column four: the last person to touch the ball before it died. Those four columns will not tell you who the best player in the league is. They will tell you where your team is losing points and how. After ten matches you own something ten years of scoring reports never produced: a sample large enough to speak the word system.

Football stopped rolling in 2026, but my data never stopped. Domestic volleyball never stopped rolling, but its data has never started. Who will be the one sitting down to fill in the first column next round?