Two hitters finish with eight kills. Both attacked twenty times. The first lost four attempts directly and kept eight in play; the second lost one and kept eleven in play. Their kill totals match. The cost of those totals does not.
These are fictional figures. We will use them to build a record that another person can check and turn into a practical decision.
Choose the question before choosing the statistic
“Who attacked better?” is still broad. Do you want to know who finished more balls, lost fewer, handled difficult sets or converted after a dig?
The question defines the sample. One table may answer the first question and lack the information needed for the others. Without recording the setting situation, it is difficult to separate routine opportunities from emergency swings.
The FIVB technical report for the 2025 VNL distinguishes in-system and out-of-system attack analysis. That distinction is useful because context belongs with the number. A team does not need to copy every elite-competition indicator to start asking a clear question.
Keep a small record that can be reconstructed
Begin with fields that let you find the play again: set, score or rally number, player, action, conditions and result. Add the video timestamp if video is available.
| Field | Example entry |
|---|---|
| Situation | Transition after a dig away from the net |
| Player and action | Outside hitter attacking from the left pin |
| Conditions | High set against a formed double block |
| Immediate result | Opponent dig; rally continues |
| Final rally result | Own team wins on a later action |
The recorded attack did not become a kill because the team eventually won. Giving the first hitter credit for every later rally win silently changes the meaning of the column.
Also define what happens when an action cannot be observed. “Unidentified” preserves uncertainty. Guessing a category creates precision that the observation did not support.
Eight kills can describe different outputs
For this example, every attack attempt receives one immediate result: kill, direct loss or continuation. A direct loss includes an attack out, into the net or blocked down to end the rally against the attacking team. Count each attempt once.
| Fictional record | Hitter A | Hitter B |
|---|---|---|
| Attempts | 20 | 20 |
| Kills | 8 | 8 |
| Direct losses | 4 | 1 |
| Continuations | 8 | 11 |
| Kill rate: kills ÷ attempts | 40% | 40% |
| Net return defined here: (kills − direct losses) ÷ attempts | 20% | 35% |
The first percentage measures how often an attempt ended in a kill. The second deducts direct losses according to the convention stated here. Do not swap their names or compare either with another provider's metric without checking its definition.
The table has not yet selected the starting hitter
Add one fact: A received many off-net sets against an organized block, while B received more balls with several attacking options available. The calculations remain correct. They do not establish that B would solve A's opportunities better.
Separate similar conditions and return to the rallies. If A repeatedly loses by attacking into a closed block, the attacking choice deserves attention. If A cannot reach the set in a useful position, the adjustment involves delivery and the availability of other options.
The number identifies a question. The sequence helps answer it.
First-swing kills and rally wins are different measures
Consider another fictional sample: twelve rallies beginning with the team receiving a valid serve. The team wins seven, but only four end on its first attack.
Rallies won while receiving: 7 ÷ 12, approximately 58.3%. First-attack kills per those receiving rallies: 4 ÷ 12, approximately 33.3%. These examples share a denominator but measure different outcomes.
If the team receives well but needs several constructions to score, the first measure can hide a first-swing finishing problem. If the initial attack improves but subsequent defensive rallies are lost, the second can hide a continuation problem.
The side-out guide develops this phase. When comparing reports, check how opponent service errors enter the sample. A denominator is a definition, not just the bottom number in a formula.
Move from the spreadsheet to a specific adjustment
Imagine that the record highlights repeated losses in one rotation. The first reaction is that the outside hitter needs to attack better. Video reveals a more precise sequence: the hitter receives a short serve, starts the approach late and meets a set that is too fast for the preparation time left.
The first intervention could be more time in the set under that condition. Keep the receiving responsibility and compare similar later opportunities. Look for the hitter reaching a position that allows a choice of contact. Points still matter, but they are not the only sign that the intervention addressed the cause.
If the encounter improves and losses continue through the same choice against the block, the next intervention changes. Now direction or an attacking tool may be the relevant subject. The record helped by separating two causes that both appeared in the broad category “attack error.”
That connects attacking with reading the game. A table ending with “the player was poor” has not completed the analysis.
Keep sample size and coding criteria visible
Two successful attempts out of three create a high percentage. A loss on the next attempt changes it substantially. Always show the number of opportunities alongside the rate, especially in short samples.
If two people record the game, compare several plays and align categories. One person's positive pass may be another person's limited opportunity. Averaging incompatible classifications does not resolve the disagreement.
Preserve conditions when assessing an adjustment. Comparing the starting lineup's organized attacks with another group's emergency balls combines several changes at once. A record can contain both, but it must allow them to be separated.
Volleyball statistics FAQ
Do I need paid software to begin?
No. A sheet with a few clear categories can answer a specific question. Software helps manage volume and video, but it does not define the categories or guarantee the quality of the observation.
Can an attacking percentage be negative?
A net measure that subtracts losses from kills can be negative when losses exceed kills. A simple kill rate cannot. Keep the name and formula beside the value so the reader knows which measure it represents.
What if there were no attempts?
Report that no attempts occurred. Division by zero does not produce a valid rate. Showing 0% would imply there were opportunities but no conversions; that is a different situation.
To deepen the game reading that turns information into setting and attacking decisions, explore Código R17.
