Volleyball and the Limits of Data: When an Empty Table Is More Honest Than a Rushed Conclusion

On my computer screen in Nagoya, the spreadsheet opens with five familiar columns: spike success rate, blocks per set, ace-to-error ratio, perfect-pass rate, and dig rate. This is the framework I have used for nearly every volleyball analysis across many years. This time, after rewatching the match and sitting down at the desk, I counted that nearly half the cells were still empty. It was not that the match lacked situations to record. The problem was that public data could not fill those cells, and internal data was beyond my access. I could write a long, confident, smoothly worded piece whose core would still be hollow. That moment of staring at an empty table forced me back to a question I had avoided for years: at what point does the absence of information itself become a finding worth publishing? In volleyball, the public data layer is far thinner than what audiences are used to in other sports. A second-division football match can yield dozens of detailed metrics per player. Volleyball mostly stops at team level. The five foundational metrics are usually published after the match, but they are merged into a team average. To know how efficiently a wing spiker attacks against a tall block, or how steadily a libero passes against heavy spin serves, the analyst must break the data down personally. That is why my work remains heavily manual. In every viewing session, I sit with a notebook and a spreadsheet open side by side. I record every serve, every blocking situation, every failed reception. When I say I watched one team for seventeen matches just to find a single gap, that is not a figure of speech. It is the only way to obtain data I can trust. But that very manual nature sets limits. Human capacity is finite. One person cannot record enough detail across an entire season in multiple competitions at once. When resources fall short, the analyst faces two choices: narrow the scope to go deep, or widen the scope and accept the holes. I choose the first, and its price is that I frequently have to say I do not know, in the face of questions readers consider simple. There is another data layer I know exists but cannot reach: the internal data of teams and federations. There, people track metrics that never go public, such as a libero's reaction time, a wing spiker's approach path, or a setter's position in each combination. Those numbers decide selection and tactics, but they are locked inside the analysis room. An outside writer like me only sees the tip of the iceberg. That information asymmetry shapes the whole way I write. I must build conclusions from what can be observed, while clearly marking the zones where I have no basis to assert. Many colleagues choose to fill the empty zones with plausible-sounding speculation. I keep the gaps intact, and turn them into part of the analysis. The five metrics and what they miss Start with spike success rate, the most cited metric. It shows what percentage of attack attempts ended in a point. It sounds clear, but it merges every kind of situation: a spike after a perfect reception cannot be compared with a spike after a scrambled high ball. A wing spiker hitting forty-five percent on a good combination foundation may be less effective than one hitting thirty-eight percent in a system where he must create his own chances. Reading only the final number means skipping the hardest part of the story. Blocks per set is the same. It counts successful blocks per set, but says nothing about the blocks that graze the ball and slow it down for the defense behind to dig. In modern volleyball, the value of a block lies in where it directs the ball, not only in whether it scores directly. A well-timed deflection can turn a dangerous attack into an easy dig for the libero. A raw block metric does not see that causal chain. The ace-to-error ratio is a metric I treat with special caution. It rewards risk-taking and punishes safety, but elite volleyball often operates on the reverse logic. A powerful serve that forces a bad reception, even without a direct ace, may be worth more than an ace followed by three consecutive errors. The ace-to-error ratio cannot measure the pressure a serve places on the opponent's reception system. Perfect-pass rate is the metric closest to truth, yet it depends on how perfect is defined. Each league and federation may use a different scale. A pass counts as perfect when the ball reaches the setter's position, but the quality of that pass, including height, spin, and speed, is not reflected. Two liberos with the same perfect-pass rate may be giving their setter entirely different levels of condition. Dig rate, the last metric, is most heavily influenced by positioning. A team defending deep will have more digging opportunities than a team pressing high, simply because more balls reach their positions. Comparing dig rates between two teams running different defensive systems is comparing two things that are not of the same nature. Context decides the value of a number If those five metrics are already thin, the problem of context makes them even harder to read. A spike success rate only means something when you know which block it was produced against. A team facing only weak opponents will post pretty numbers that do not reflect real ability. In a short season, just a few matches against strong opponents can pull an average down considerably, or the reverse. I learned this from a habit of verification. When a model produces a beautiful result, the first thing I do is try to break it. Before publishing any model, I try to break it first. In volleyball, the simplest way to break it is to split the data by opponent strength. If a wing spiker only performs highly against teams outside the top eight, his numbers cannot be used to predict anything in the knockout rounds. Strong and weak teams also operate at different tempos. Strong teams often end rallies earlier, meaning they have fewer long rallies. That affects every defensive metric. A libero on a strong team may have fewer digging opportunities than a counterpart on a weak team, not because he is worse, but because the block in front of him works better. Reading a metric while ignoring system context is reading half the story. The reception system: the biggest blind spot Among all components of volleyball, the reception system is where public data is weakest relative to its true importance. Everything in attack begins here. Reception quality determines how many options the setter has, and the number of options determines how much the opposing block must guess. A team can reach a sixty percent perfect-pass rate, but if the remaining forty percent are severe failures, the attack system collapses at precisely the decisive moments. Conversely, a team reaching only forty-five percent perfect, with the rest merely acceptable, may run more smoothly. The average hides the distribution, and in volleyball the distribution matters more than the mean. I track the reception system by recording the position of each pass, not just its outcome. A pass that arrives at the right spot but at low height forces the setter to handle it quickly, reducing the ability to combine with a wing spiker. Such details appear in no public statistical table. They exist only in the notebook of the person watching. When I doubt a number, I watch more matches to verify. Once I read a model claiming that serving pressure raised defensive efficiency by about thirty percent. I doubted that rate, so I watched fifteen more matches before believing it. The verification did not fully refute the model, but it showed the number depended heavily on which team was serving. Context, once again, decided the meaning of the data. The setter: the decision-maker with no public data If reception is a blind spot, the setter is a fully dark zone. No public metric measures the quality of a setter's decisions. People count successful sets and scoring combinations, but they cannot measure whether the setter chose the right option in a specific situation. A great setter is not the one who delivers the most scoring sets, but the one who reads the opposing block and distributes the ball to maximize the team's chances. Sometimes the right option is the one that does not score, because it pulls the block out of position for the next play. Those decisions are invisible in every statistical table. To assess a setter, I must rewatch each rally and note his choices in relation to the opposing block's position. It is time-consuming work, and it yields no tidy number to cite. But it is the only way to understand why a team attacks effectively, or why a team with enough good spikers still stalls. In women's volleyball, where combination speed is higher and the margin for error smaller, the setter's role is even harder to measure. A fast combination at the number three position demands timing accurate to fractions of a second. No public metric captures that precision. The analyst can only describe, compare by eye, and admit his limits. Block and defense: a coordination problem Block and defense are two halves of one system, but they are often statistically separated. A good block makes the defense look better, and vice versa. Because the metrics are recorded separately, readers easily credit individuals when the result is really coordination. What I look for when reviewing defensive plays is not who dug the ball, but who moved to create the digging chance. A player's position before the ball is spiked matters more than the dig that follows. A libero in the right spot will dig balls that look easy; a libero in the wrong spot will dive in vain. Statistics record only the action, not the position. The gap a player leaves behind when he rises to block is where the match is decided. I watched seventeen matches just to find the gap a player leaves behind. When a blocker rises high, he opens a zone behind his back that the defense must cover. If that coverage does not arrive in time, the opponent only needs to spike into it. No metric measures the gap, only the human eye and the notebook. Repetition as evidence When data is insufficient to prove, repetition becomes substitute evidence. A situation occurring once is random. Occurring three times is notable. Occurring nine times, it has become a feature of the system. In volleyball, I count how often a pattern appears before turning it into a judgment. If a team repeatedly exposes a gap at the number six position every time a blocker rises, that is no longer an accident. It is a structural flaw, and it will be exploited in the most important matches. That counting demands patience. It does not produce fast results. But it produces results that can withstand challenge. When I publish a judgment based on repetition, I can point to each occurrence, with timing and context. That is a kind of evidence far harder to refute than an average. Sample size: the trap of a short season Volleyball has a feature that makes data analysis harder than in many other sports: fewer rallies per match, and fewer matches per season. A five-set volleyball match may contain only about two hundred rallies. Compared with thousands of situations in a football match, that is a small sample. A small sample means every metric fluctuates widely. A wing spiker hitting fifty percent over three matches may hit only thirty percent over the next three, and both numbers fall within normal random range. Anyone reading volleyball data must always ask whether the difference they see is a real signal or just noise. In a month, I can watch up to sixty-four matches, and every dead-ball situation is recorded by me. One month, sixty-four matches, and every dead-ball situation recorded by me. Yet even with that volume, I must stay cautious when drawing conclusions about a specific player. Sixty-four matches is a lot for one person, but still small against the demands of inferential statistics. Nine powerful serves were repeated, until the two-zero score became a warning. In volleyball, patterns that repeat in dead-ball situations are often more reliable signals than averages, because they occur in clearly structured situations. That is why I built a separate database for dead balls, apart from open-play data. The Olympic cycle and the pressure of compressed time At national-team level, the data problem becomes more urgent. The Olympic cycle lasts four years, but a national team's official matches within that cycle are very few. Competition windows are compressed into a few major tournaments, and between them lie long breaks with no official matches. That means every assessment of a national team rests on a tiny sample. A continental championship may have only five to seven matches. Olympic qualification may be packed into a single week. The analyst lacks enough data to separate signal from noise, yet must still offer a judgment. That pressure pushes many writers toward conclusions stronger than the data permits. A lopsided win becomes proof of a new system; a loss becomes a sign of crisis. Both are leaps the data cannot support. My way of handling it is to return to what has a more stable structure: dead-ball patterns, lineup changes by situation, and how a team reacts when trailing. Those repeat often enough to trust, even when the total number of matches is small. A national team may play only five matches in a tournament, but if in all five they handle serves the same wrong way, that is a signal, not noise. Professional standards across markets There is another trap I try to avoid: imposing one market's way of reading a match onto another. Japanese volleyball and volleyball in other nations operate by different standards, both in tempo and in system organization. A metric considered normal in one place may be abnormal in another. When analyzing a team, I try first to understand the logic they were trained under. Some volleyball cultures emphasize risky serving and accept errors in exchange for pressure. Others emphasize stability and limiting mistakes. Judging a team by another culture's standards yields skewed conclusions, however accurate the arithmetic. This also holds true at league level. A mid-table team in a strong league may be stronger than a top team in a weak league, but public metrics do not tell us that. To compare, the analyst must adjust to each league's standards, and that adjustment always carries a degree of subjective judgment I must acknowledge. The blind spot of certainty The irony is that insufficient information often leads not to caution, but to overconfidence. When data is thin, the gaps get filled with narrative. A winning team is assigned some spirit; a player performing well for two matches is called the discovery of the season. Such stories sound plausible precisely because they fill the gaps the data leaves. I hold that the analyst has a duty to keep the gap intact rather than fill it. A table with empty cells is an honest table. A conclusion built on those empty cells without acknowledging them is a dangerous conclusion, because it creates a feeling of certainty where there is no basis. There is a paradox in how the public receives sports analysis. The more decisive the writer's conclusions, the more attention they get. The more a writer admits limits, the more they are seen as indecisive. That incentive structure pushes an entire industry toward claims stronger than the evidence. In volleyball, where data is already thin, the consequences are heavier. But I must also guard against my own stance. Excessive caution can become a form of avoidance. If I refuse every conclusion because the data is not yet enough, I contribute nothing beyond pointing out that everything is uncertain. Once I have watched seventeen matches and seen a pattern repeat, I have a duty to speak, with a corresponding level of confidence. Honesty lies in quantifying the degree of certainty, not in refusing every assertion. So the question I ask myself is not whether the data is enough, but how much I am willing to bet on this conclusion. For a dead-ball pattern repeated across many matches, I am willing to bet high. For a single metric in one match, I bet almost nothing. Between those two extremes lies the entire space of the analyst's craft. What to verify in the next match As the season continues, I will track whether the dead-ball patterns I recorded keep repeating, and whether teams' reception systems hold their stability across matches against stronger opponents. If a team keeps its serve-handling approach and still wins, that is a signal their system is genuinely solid, not merely lucky with an easy schedule. The question I carry into the next match is simple: when data is insufficient, will I have the courage to say I do not know, or will I once again fill the gap with a plausible-sounding story?