Volleyball Data: When the Sample Size Is Too Small, Conclusions Must Be Suspended

On a volleyball forum, a box score is shared with a short caption: "48% attack, and people say she has declined." The score reads 22 points, 45 attack attempts, a 48.9% success rate. Nobody in the thread mentions the nine attack errors and five times blocked sitting on the same row. Take 22 points, subtract nine errors and five blocks, and the numerator falls to 8 out of 45 attempts — an actual attack efficiency of 17.8%. Same match, same player, two numbers nearly three times apart, and only one of them reflects attacking value.

I retell this not to score points off a forum. I retell it because it is a symptom of a habit that has sunk deep into how volleyball is read across Southeast Asia: take the easiest number, the most legible number, and hang a conclusion on it. Success rate becomes a label. The denominator, the thing that decides everything, disappears from the conversation. Over recent months, working through Data Volley exports — the technical scouting and statistics software common across professional volleyball — I have come to see that most arguments online are not about who is right and who is wrong. They are about nobody reading the same definition.

Every dataset tells a story; we simply are not patient enough to listen. The trouble here is that people are listening to the wrong chapter.

Two divisors, two conclusions

Spike success rate and spike efficiency are two different divisions, and confusing them is the industry's most common distortion. Success rate divides spike points by total attempts. Efficiency subtracts attack errors and times blocked from spike points before dividing by total attempts. The second calculation punishes carelessness; the first does not. An attacker can post a 50% success rate with a 10% efficiency if her errors and blocks roughly match the points she scores.

The worry is not that fans do not know the formula. The worry is that match reports, season reviews, and even "official" statistical tables routinely pick whichever division flatters the story they want to tell. When a statistical table does not state its definitions, it is not data — it is an advertisement wearing numbers as a costume. Readers are not wrong to trust a box score. Publishers are the ones who owe them a definition.

Volleyball carries another layer that football does not replicate at the same level: every metric is downstream of the reception system. Perfect-pass rate — the share of first contacts delivered to the ideal spot so the setter can open the full attacking menu — determines what kind of ball an attacker receives. An outside hitter posting 45% efficiency behind a 60% perfect-pass rate and one posting 38% behind a 40% rate are players at different tiers, yet the box score seats them side by side. The comparison is meaningless without the first-contact number.

At national-team level the problem compounds. No opponent-strength adjustment standard is widely applied across Southeast Asian competitions, where the gap between the leading group and the bottom of the table can span two tiers. Twenty points against a side ranked three places lower is not worth fifteen points against a peer. Box scores do not say so, which is why I keep a separate notes column beside every dataset I handle.

The discipline of not rushing to conclude

Based on my own experience watching matches in the Thai league and across regional competitions, I built a nine-layer process for reading any volleyball match before I permit myself to write a single conclusion. The nine layers are not ritual. They are fencing against my own impatience.

Layer one: tactics and technique. Before touching a number, I establish what system the team runs, how the rotations are arranged, and — most importantly — where the primary attacker receives the ball. A team can win three sets in a row purely because the opponent served badly, and the box score cannot distinguish that from a win built on a varied attacking system. The concept of a "stuck rotation" — a side repeatedly unable to side out while the opponent stacks points — only surfaces when you track the point sequence rotation by rotation. It never surfaces in a final aggregate.

Layer two: data. This is where I spend the most time and where the reading is laziest. Five core metrics come first: spike rate and efficiency, blocks per set, the ratio of aces to service errors, perfect-pass rate, and dig rate. Every metric must arrive with three companions: the definition in use, the sample size, and opponent-strength adjustment. Aces without service errors is half a truth. An aggressive serving team can take eight aces and concede fourteen service errors — a net negative. The box score prints only the first half.

Numbers do not lie, but they know how to hide the truth. Their favourite hiding place is letting you choose the denominator yourself.

Layer three: competition system and schedule. The same form guide means something entirely different in an Olympic year than in a mid-cycle adjustment year. Olympic qualification turns every friendly into a deliberate tactical rehearsal; a mid-cycle Volleyball Nations League is where teams test rosters and bank world-ranking points. The VNL is staged annually by the FIVB, volleyball's global governing body, and carries both commercial weight and ranking points, so the value of a win there shifts year to year. Ignore that context and a rehearsal reads like a decider.

Schedule density is the second variable here. A national-team player who finishes a domestic season, flies home, joins camp for four days, then enters a continental championship carries a different physical baseline from one who rested two weeks. No box score reflects that. Only the calendar does.

Layer four: landscape and positioning. Placing a team on the international ladder demands at least one comparison opponent. Title contender, medal contender, quarterfinal-level, second tier — these cannot be assigned in a vacuum. At the same time I look at the resource structure: bench depth, youth-development output, and the support a domestic league provides. A team strong in the starting six but thin on the bench is a team at risk of collapse under a compressed schedule.

Talent-flow signals matter just as much. Outbound players, naturalised players, and the risk of a talent cliff when one gifted generation crosses its peak simultaneously — these three decide a national team's three-to-five-year trajectory far more than any single result.

Layer five: rules and governance. This is where volleyball media slips most often. Insinuating a compliance problem from the mere existence of an article is a methodological failure. Without a specific decision, a specific rule, and a specific body with jurisdiction, there is no compliance issue to discuss. The International Transfer Certificate — the ITC — is the mandatory document governing player moves between federations, and any cross-border transfer story should start there before reaching for anything else.

Layer six: team building and personnel. Age curves and injury history form the backbone. A 30-year-old attacker with a knee history, carrying both a club season and a national-team season, carries a different decline probability from a 24-year-old who is rotated. Coaching-change signals — results pressure, contract cycles, the opinion arc from praise to condemnation — are only legible when you know the team's state: stable, transitional, or turbulent.

Layer seven: the risk surface. I sort risk into six groups: competitive, personnel, schedule, rules, public opinion, and systemic. But there is a seventh I always place last at the highest level: analytical risk. That is the danger of a document presented neatly, correctly formatted, with every heading in place, containing no findings — and a downstream reader mistaking it for a real assessment. Formal polish can generate an illusion of substance, and that illusion is more dangerous than a wrong number.

Every statistical table is a forest; I am only the one reading animal tracks. A track does not mean the animal is still there. It only means the animal once walked through.

Layer eight: public narrative. A Southeast Asian women's national team winning three straight matches immediately generates a story about spirit, identity, and golden ambition. The story may be true, but it is unverified. The sustainability test has two questions: does it rest on underlying fundamentals, and does it survive a larger sample? A narrative sustained only by emotion has a short life, and when it dies it drags public trust in the entire analytical field down with it.

Layer nine: industry transmission. Finally, I trace an event through the chain from upstream to downstream: youth development and talent supply, domestic leagues and national teams, broadcasting, commerce, and derivative industries. A change in foreign-player policy in one domestic league propagates to domestic youngsters' minutes, then to national-team strength, then to broadcast-rights value. The chain is long, and most reports photograph exactly one link.

Correlation is not causation

There is one error I find myself most prone to: seeing two things move together and concluding one caused the other. Teams win matches in which they block a lot. Therefore good blocking wins matches. But it may be that leading on the scoreboard forces opponents into harder shots, making blocks an effect of the game state rather than its cause. I once saw a season dataset in which the league's most prolific blocking team finished seventh. There is no paradox there. There is only an omitted variable standing behind both.

Data does not make decisions; it only kills doubts. But it kills only the doubts you are willing to define precisely. Vague doubt is beyond its reach.

This leads to a counter-intuitive conclusion I learned from my own mistakes. An empty analysis — one halted because there was not enough data to conclude — is not a failed analysis. It is evidence that the process is working. The greatest value of an analytical system lies not in how many conclusions it produces but in how many unsupported conclusions it blocks. A report willing to write "insufficient information to assess" is more trustworthy than one stuffed with judgements and not a single sourced figure.

That holds especially true in the transfer market. Noise drowns signal. Attackers such as Tran Thi Thanh Thuy or Chatchu-on Moksri are names repeated constantly yet rarely read correctly, because their value sits in metrics the box score never prints: perfect-pass rate on difficult balls, backcourt dig rate, and points scored in broken-play situations — three numbers agents rarely cite because they sell no glamour.

I do not write to prove I am right; I write to find where I was wrong. Every time an analysis is suspended for lack of data, I have saved myself a wrong conclusion I would have had to retract six months later.

A signal for the next cycle

Over the coming cycle I will track three things and date them precisely, not by week. First, the perfect-pass rate of Southeast Asian national teams against higher-ranked opponents, because that number sits beneath almost every attacking table. Second, the net ratio of aces to service errors, because half of that calculation is routinely discarded. Third, the number of rest days between the domestic season and national-team camp, because that variable appears in no statistical table yet decides a whole tournament's efficiency.

If those indicators are not thick enough to conclude from, I will write exactly four words into the report: insufficient data. Fans do not need a destination; they need a map. And an honest map is one that dares to draw the regions it has not yet surveyed.