When the Data Table Comes Back Empty: Vietnamese Volleyball and the Integrity Check

When the Data Table Comes Back Empty: Vietnamese Volleyball and the Integrity Check

11:47 p.m., August 12. The match in Nha Trang had finished forty minutes earlier. The stands were dark, and a half-finished mop streak still marked the corner of court number four. I opened my analysis system, pasted the article link into the input field, and pressed run. Nineteen seconds later, the screen returned a table. The "Information Points" column was empty. The "Entities Involved" column was empty. The "Article Title" field read N/A. The "Article Source" field read N/A. No team name, no person, no score, no date.

A nine-dimensional analysis table was generated with every heading in place, every sub-table drawn, every assessment cell formatted. And every content cell carried the same sentence: insufficient information to assess.

This was the fourth time in two years I had seen this happen. The first time, I assumed I had pasted the wrong link. The second time, I sat down and audited my own scraping code line by line. The third time, I started taking notes on the phenomenon instead of fixing it. This time, I decided to write about it.

Data never lies, but it knows how to hide. The problem is that most of the time it does not hide where we assume it hides. It hides at the collection layer — the layer nobody in Vietnamese volleyball wants to discuss, because discussing it means admitting it barely exists.

***

A men's or women's match in Vietnam's national volleyball championship, on average, runs between 90 and 135 minutes for a three-set match, or 130 to 180 minutes for a five-set match. In that window, the two teams touch the ball somewhere between 3,200 and 4,100 times. The number of complete rallies — from the first serve until the ball hits the floor or a fault is called — ranges from 165 to 210. Each of those rallies contains at least twelve independently recordable events: serve position, serve velocity, ball landing point, first passer, quality of the first pass, setter, set location, attacker, attack direction, blocker, back-row defender, and the final outcome.

That means a volleyball match contains roughly 2,000 to 2,500 independently recordable events.

The data actually recorded and published in Vietnam, across the entire national league system in the most recent season I tracked: set scores, match score, individual points, direct service aces, successful blocks, and total errors. That is nine data fields covering roughly 2,200 events. A coverage rate below 0.5 percent.

I am not saying this to accuse anyone. No organiser is obliged to record the first pass of a libero in the seventy-fourth minute of a group-stage match. But I am saying it to raise a professional question: if 99.5 percent of what happens on the floor is never recorded, what exactly are our tactical judgements about Vietnamese volleyball built on?

The honest answer is: collective memory. And collective memory, like any uninstrumented tool, tends to remember what was beautiful and forget what was accurate.

I have tested this over years of watching matches in Nha Trang and across the national system. When I ask ten spectators whether a team played well, I receive on average seven answers built on the two or three most spectacular rallies that team produced in the set. Nobody remembers the thirty-fourth rally, where the first pass landed half a metre off target and the entire attack system collapsed in four seconds. But those are the rallies that decide the score.

The night Germany collapsed, I learned to test my own assumptions. On June 27, 2026, I sat in a university dormitory and stayed up until one in the morning watching a football match. My spreadsheet tracked the running distance of the midfielders, and the figure came in almost 1.4 kilometres per match below the standard set four years earlier. I shouted that the team would lose before the second half began. They did lose. But the lesson I carried away was not "I was right." The lesson was: I was only right because I had a table of numbers to check my assumption against.

In Vietnamese volleyball, we usually do not have that table. And so we never get the chance to find out where we are wrong.

***

Consider a comparison. In football, a metric such as PPDA — passes allowed per defensive action — took about fifteen years to travel from the laboratory to the commentary box. It is not a perfect metric, but it lets people argue about something specific: whether a team presses high or low, and whether that pressing works.

Volleyball has an equivalent, arguably more sophisticated set of metrics, and it has existed for a long time in international competition systems. Let me list the metrics that any analysis unit inside an Asian federation uses daily.

Perfect-pass rate — the share of serve receptions delivered to the exact spot that allows the setter to run the full tactical menu, usually defined as a ball landing within one and a half to two metres of the net, in the central corridor. This is the root metric of every attack system.

Sideout rate — the share of rallies a team wins after the opponent's serve successfully crosses the net. This measures a team's strength from a passive position.

Break-point rate — the share of rallies a team wins while serving. This measures its ability to apply pressure from an active position.

Attack efficiency, calculated as points scored minus attack errors divided by total attack attempts. This is the most misunderstood metric in Vietnam, and I will return to it.

Blocks per set, alongside a rarely used secondary metric: the number of blocks that touch the ball and slow it down without scoring directly. That secondary metric matters more than the headline one in many defensive systems.

Back-row dig rate — successful saves divided by total opponent attacks entering the back-court zone.

Out-of-system attack share — the percentage of attacks executed after a substandard first pass. This measures how much a team depends on individual ability rather than system.

And the last metric, the one I consider most important and least discussed: point differential by rotation. In volleyball, the lineup rotates with the service order, and each rotation presents a different configuration of front-row players. Some rotations have three attackers at the net. Some have only two, and that is a structural weakness. A team can be strong overall yet lose points repeatedly in exactly one rotation. If you do not record point differential by rotation, you will never see that gap.

I spent three seasons recording point differential by rotation for matches in Vietnam that I could watch live or on recorded footage. The number of matches I recorded in sufficient detail was forty-one across three seasons. That is a small sample, and I will state its limits clearly in the data-credibility section. But across those forty-one matches, one pattern appeared so repeatedly that I was compelled to write it down.

***

The pattern is this. In most of the matches I charted, the point differential between rotations within a single team was larger than the average point differential between the winning team and the losing team across the whole match.

Put differently: the gap between a team's best rotation and its worst rotation was frequently larger than the gap between that team and its opponent.

This finding has direct consequences for how we coach and how we commentate. If the cause of a defeat lies in rotation structure, then every explanation invoking "spirit" or "nerve" is aiming at the wrong target. And in Vietnam, those two words appear in roughly seventy percent of the post-defeat commentary pieces I read.

I do not deny the role of competitive psychology. I deny the use of psychology as a default explanation when there is no data to rule out other causes.

Picture a specific situation I recorded. Team A led 18-15 in the fourth set. In the next rotation, Team A had only two attackers in the front row: one middle blocker and one outside hitter. Team A's setter was forced to distribute across two options, and the opponent needed only one blocker focused on the outside hitter to read the intent. Over the next seven rallies, Team A lost five straight points and let the opponent move ahead 20-18. The set finished 25-22 for Team B.

When I read the post-match commentary, the phrase "psychological collapse" appeared fourteen times. Across the seven rallies I had charted in detail, Team A committed no psychological error. They produced six first passes that met the standard. The problem lay in a structural fact: with only two front-row attackers, a team's scoring probability drops to a level that no mental skill can compensate for.

That is something a data table can show you in thirty seconds, and something an emotional commentary will never see.

***

Now to the most misunderstood metric: attack efficiency.

In Vietnam, when people discuss an attacker, they usually discuss total points. An attacker who scores twenty-two points in a five-set match is considered to have played well. An attacker who scores fourteen points is considered to have been quiet.

That reading ignores three variables that can reverse the conclusion entirely.

The first is the number of attempts. Twenty-two points on sixty attempts is a 36.7 percent rate. Fourteen points on twenty-five attempts is a 56 percent rate. The second attacker performed almost twenty percentage points better, while receiving less than half the sets.

The second is errors. Attack efficiency in professional volleyball is calculated as points minus errors, divided by total attempts. An attacker who scores twenty-two points but commits twelve attack errors and is blocked seven times has a true efficiency of five percent. That number appears in no Vietnamese match report.

The third and most important variable is the quality of the first pass before each attack. An attacker hitting off a perfect first pass has a materially higher scoring probability than an attacker forced to handle an out-of-system ball, one delivered to a position that prevents the setter from running the playbook.

In Vietnam, the out-of-system attack share for many teams in the matches I charted fell between 35 and 45 percent. That means nearly half of all attacks occur under structurally unfavourable conditions. Attributing low efficiency to individual ability in that context is a serious methodological error.

And here I want to be blunt about a trend spreading through how Vietnamese players are evaluated in recent years: using points as the sole measure of an attacker's value has produced a system of perverse incentives. It rewards players for taking volume, not for selecting good sets.

Transfers are not addition, they are the algorithm of greed. A club pays for last season's point totals, not for this season's efficiency inside a system. And when that attacker arrives at a new club, with a new setter and a new reception system, their numbers usually fall. Nobody can explain why, because nobody has the old team's first-pass data for comparison.

***

Let me return to the empty pipeline, because I believe it is an exact metaphor for the current state of Vietnamese volleyball.

When my analysis system returned an empty table, it did not raise an error. It did not crash. It returned a structurally complete table with every heading and sub-heading in place, writing the phrase "insufficient information to assess" into each content cell.

A hasty reader would look at that table and assume the analysis had been performed. The table had a table of contents. The table had sub-tables. The table had a conclusions section. Only the content was empty.

To me, that is precisely the state of most tactical commentary and evaluation in Vietnamese volleyball today. We have the full table of contents: dedicated columns, guest analysts, pre-match previews, post-match reviews. We have all the major headings. The problem is in the content cells.

The root cause is not the competence of the writers. The root cause is at the collection layer. No system records the first pass. No system records point differential by rotation. Nobody records the count of touch-blocks. And because nothing exists at the lower layer, the upper layer is forced to fill the void with the only thing left: feeling.

On July 15, I spent an entire afternoon cross-checking three different public data sources for the same match in the national league system. The result: the leading attacker's point total matched across all three sources. Successful blocks diverged between two sources. Total errors diverged across all three, with the largest discrepancy at seven errors in a single match. The number of rallies in the match was recorded by none of them.

Seven errors of divergence in a five-set match is an error margin large enough to flip the conclusion about which team played more disciplined volleyball. And this is data at its simplest layer, the layer everyone assumes is accurate.

***

I have set myself a working rule for years now, and I call it the three-hundred-character rule.

The rule is this: before making any judgement about a match, I must hold at least three hundred characters of raw data, with a clear source, a date, and the property of being re-verifiable by someone else. Three hundred characters, not three hundred complex figures. Just twelve honest lines of notes on twelve consecutive rallies.

That sounds like a low bar. But I challenge you to find three hundred characters of verifiable raw data about any national championship semi-final in Vietnam from the last three seasons, beyond the score and the list of scorers.

The rule sounds extreme. It is no more extreme than a doctor refusing to diagnose before seeing test results. Before you burn the tactics, check your data source.

I do not believe in instinct. I believe in the moment instinct becomes digitised. The instinct of a seasoned coach is an uncoded data signal. It has value. But it cannot be transmitted, cannot be verified, and cannot be improved upon by a successor. Only when it is written down does it become an asset of a volleyball culture.

In Vietnam, we have a great many excellent instincts and very few recorded assets.

***

At this point I must address the limits of what I have just presented, because otherwise I would commit exactly the error I am criticising.

Forty-one matches across three seasons is a small sample. It is not enough for statistical conclusions at league level. It is skewed by the fact that I only charted matches I could watch, and the matches I could watch were mostly those of teams in the south and those with television coverage. That is a clear selection bias.

Second, my manual charting of first passes introduces an irreducible subjective element. I graded first-pass quality on a three-point scale. Another charter would grade differently. I re-tested myself on ten matches three months later and the agreement rate was about eighty percent. That twenty percent of divergence is the price of having no automated charting system.

Third, and most important: correlation is not causation. Finding a link between a weak rotation and a run of conceded points does not mean that rotation is the sole cause. It may be the consequence of a deeper cause — for instance, a middle blocker's stamina dropping after the third set, leading the setter to avoid distributing to that position, leading the opponent to concentrate blocking on the two remaining options. In that case, the rotation is where the symptom surfaces, not where the cause lives.

This is why I always write "my model gives a probability" rather than "certainly." A model built on forty-one matches can be right about the trend and wrong about the individual case. And in sport, the individual case is what gets written into history.

***

There is a dimension data never touches, and I want to spend this section on it, because ignoring it makes every analysis inhuman.

Among the forty-one matches I charted, there is one I remember in precise detail. It was a match the home team lost in five sets, and in the fifth set their starting setter cramped up in the seventeenth minute. I have the full data for that set. After the starting setter left the floor, the home team's perfect-pass rate did not fall much. But their out-of-system attack share rose from 38 percent to 61 percent, and overall attack efficiency dropped from 44 percent to 19 percent.

My data table explains what happened on the floor in those ten minutes. It does not explain why that young woman cramped. It does not tell me how many minutes she had played in the preceding twenty days, how many kilometres she had travelled, how many hours she had slept.

This is where I want to talk about load management, a concept currently being romanticised in Vietnamese sport in exactly the way it was romanticised elsewhere.

Load management, properly done, is a medical tool. When discussed in the media, it frequently becomes a polite phrase for resting a player in one match so she can play another with higher commercial value.

In Vietnamese volleyball the problem is a degree worse, because the calendar is not designed to protect players. A female national-team player between the ages of twenty-two and twenty-seven, in a peak competitive year, takes part simultaneously in: the national championship with her club, the national cup, youth or invitational tournaments, national-team training camps, regional competitions, continental competitions, and, when the opportunity arises, a season abroad.

I have tried to count the actual match days of a Vietnamese female national-team player in a year with a full continental calendar, based on what I could chart. The figure I reached fell between ninety and one hundred and ten days with official matches. That excludes training sessions, heavy training sessions, and recovery sessions.

Under those conditions, an injury is not an accident. It is a probability outcome.

And when a player returns from injury, we get an opportunity we almost always waste: the opportunity to measure. The return phase is the only phase in which we can observe a player before and after a rupture point, and therefore it is the best natural laboratory sport possesses. But to exploit it, we need pre-injury data. And in Vietnam, pre-injury data usually does not exist.

When the stadium empties, the numbers begin to speak. I learned that in a summer when world sport stopped, when I compared more than four hundred matches with crowds against nearly one hundred without and found a marked decline in home advantage. I wrote a twenty-page report and was rejected by a sports outlet on the grounds that it was too academic. I published it myself. That piece taught me one thing: when the cheering stops, you can hear the true operating rhythm of a team.

In Vietnamese volleyball, the same approach could be applied to matches played without spectators, matches in empty arenas, or sets from which the crowd has already left. Those silences are data. We simply do not record them.

***

Fans are not a variable. They are a weight.

I say this often and I am misunderstood every time. Many people assume I am saying that spectators matter. I am not saying that, because it is obvious. I am saying that spectators change the weights of every other variable in the model.

In an arena holding four thousand people, noise changes how serves are struck. Pressure from the stands changes how a young setter chooses distribution options. The presence of a home crowd changes a team's tolerance threshold during a run of conceded points.

Therefore any predictive model for volleyball built in Vietnam must carry a crowd coefficient, and that coefficient must be calibrated separately for each arena. I have tried to do this for the three arenas I follow most closely. Preliminary results suggest home advantage in Vietnamese volleyball is not uniform, and that it depends more on the away team's travel distance than on attendance figures.

That is a counter-intuitive result, and I do not trust it fully because my sample is too small. But it illustrates a principle: when there is no data, every assumption looks correct. When there is a little data, we discover that nearly every assumption needs adjustment.

***

Now to the transmission chain of Vietnamese volleyball, viewed along the data axis.

At the upstream layer, youth development, data is close to zero. A provincial talent class may hold thirty trainees aged twelve to fifteen. Nobody measures height monthly, nobody measures jump capacity after each training cycle, nobody records each child's growth velocity. When selection moves up to provincial level, the decision rests on the coach's observation. That observation may be correct. But it leaves no trace, and therefore it cannot learn.

At the midstream layer, the professional league and the national team, data stops at basic statistics. This is the layer with the most resources and also the layer capable of changing fastest, because adding two or three properly trained charters per match would create a valuable dataset within a single season.

At the downstream layer, broadcasting and derivative markets, we see the clearest consequences. Because there is no data in the two layers above, broadcasters are forced to sell emotion. And once broadcasters sell emotion, audiences buy emotion, and the loop feeds itself.

Remarkably, this chain can be reversed. If one broadcaster decides to display perfect-pass rate on screen at the end of each set, audiences will start paying attention to it. When audiences pay attention, clubs must record it. When clubs record it, coaches must use it. Demand from the lower layer can generate supply in the upper layer.

This is the shortest and cheapest route to upgrading Vietnam's volleyball data infrastructure, and it requires no major technology investment. It requires one editorial decision.

***

I want to spend a section on risk, because analysis that omits risk is incomplete analysis.

The biggest competitive risk facing Vietnamese volleyball in the current cycle is not a specific opponent. It is a gap in improvement velocity. Volleyball nations in the region are improving on the basis of data: they adjust reception systems per opponent, they change distribution according to the opponent's rotations, they calculate in advance how many risky serves are acceptable within a set. When one side improves on a data loop and the other on an inspiration loop, the gap opens up non-linearly.

Personnel risk worries me most, because it concerns specific people. A leading attacker who must carry between 45 and 55 percent of her team's attack volume across an entire season is a fragile structure. If she is injured, the team loses not an attacker but an entire distribution system built around her over years. Building a team around one individual is a tactical decision that can be right in the short term and is always wrong in the long term.

Schedule risk has been covered above. Rules risk, at the level of Vietnamese volleyball, mainly concerns foreign-player quotas and registration windows. A small change in the number of foreign players permitted on court can restructure an entire league within a season, and no model can predict that impact because there is no historical data to calibrate against.

Public-opinion risk is the most underrated. When a team loses and there is no data to explain it, public opinion manufactures its own explanation. That explanation usually targets an individual: a setter criticised for poor distribution, when in reality she had only two options because her team's first pass permitted nothing more. And when a young player is publicly criticised for a systemic problem, we lose her.

Systemic risk is the largest and least discussed: a volleyball culture can produce outstanding individuals for decades without ever accumulating knowledge. Each generation of coaches starts over, relearning lessons the previous generation learned, and forgetting lessons the previous generation forgot. This is a kind of poverty invisible to the naked eye, and it only becomes visible when you compare the improvement velocity of two volleyball cultures over the same period.

***

I want to close by returning to that empty table on my screen at 11:47 p.m.

In that table, the "Recommended remediation" section contained four lines. I re-read them and realised that if you replace the word "article" with "match" and the word "pipeline" with "charting system," those four lines become a complete roadmap for Vietnamese volleyball.

Line one: re-fetch the source and confirm non-trivial content, with a minimum threshold of three hundred characters. For us, that means three hundred characters of raw data per match.

Line two: re-run the extraction and confirm at least three atomic sourced facts. For us, that means three verifiable events per charted rally.

Line three: confirm at least one entity is identified. For us, that means the name of the first passer in every rally.

Line four: if the source is genuinely inaccessible, supply the raw text directly. For us, that means publishing raw match data so anyone can re-analyse it and disagree.

The season is long, the data is cold, and patience is the only measure.

I do not expect a national championship to have a complete data system within one season. I expect something far smaller: that next season, in at least one match in Vietnam, somebody publishes the first-pass totals for both teams. One match. Twenty numbers.

If that happens, I will be the first to re-run my model and see whether ten years of my assumptions hold.

And if they do not hold, I will write about it. Because that is the one part of the job a data person is not permitted to skip.

***

Source note appendix

The personal records used in this article come from the author's match-charting notebooks, covering the period from the 2026 season to the 2026 season, comprising forty-one matches in the Vietnamese national volleyball league system and regional competitions involving Vietnamese teams.

The figure for recordable events within a volleyball match was calculated using the standard charting framework of international volleyball statistics, with twelve data fields per rally.

The estimate of match days for a female national-team player in a single year is a personal estimate based on published calendars of national, regional, and continental competitions during the survey period, and excludes unpublicised friendly matches.

This article makes no prediction about the outcome of any specific match. Every conclusion here is bound by the sample limits stated in the data-credibility section.

Player names, club names, and competition names mentioned in this article serve only to illustrate the general context of Vietnamese volleyball and are not attached to any personal assessment of the competence of any individual or organisation.