Possession statistics answer a question almost nobody actually asked

What the number counts

Possession share records the proportion of active playing time during which one side had control of the ball, aggregated across the whole match. That is a well-defined quantity, and the measurement of it is not seriously in dispute once the counting convention has been fixed. The dispute is entirely about what it implies, because time on the ball and progress towards the goal are not the same thing.

A side circulating the ball across its own defensive line accumulates the statistic at exactly the same rate as one carving an opponent open. Any conclusion about dominance therefore rests on an assumption the statistic itself does not contain.

Why the opponent controls the number too

A team that defends in a deep, compact block is choosing to surrender the ball in areas where holding it achieves very little. That choice mechanically raises the opponent share, so the figure partly measures a decision made by the side that appears to be losing the contest. Reading the number as a report on one team is therefore reading a joint product as though it had a single author.

This is why identical shares can describe a side suffocating an opponent and a side being invited to pass sideways indefinitely. The statistic cannot distinguish them, and nothing inside the number signals which situation produced it.

Where the ball was held

Location changes the meaning entirely, since possession inside the final third carries a different risk and reward profile from possession between the centre-backs. Splitting the figure by zone recovers some of the lost information and immediately shows that two equal shares can be built very differently. Sequence length matters as well, because many short possessions ending in turnovers indicate something quite different from fewer sustained ones.

Both refinements are available in modern data collection and both are routinely discarded when a single headline figure is displayed. The headline survives because it fits on a graphic, which is a design constraint rather than an analytical judgement.

The historical argument that muddied everything

For a period, sides built around retaining the ball were unusually successful, and the correlation was widely read as a causal endorsement of retention. The reading skipped over the fact that those sides were also exceptionally good at the things that make retention possible in the first place. Technical quality, pressing intensity and positional discipline all produce possession as a by-product while independently producing results.

Attributing the results to the by-product rather than to the underlying qualities is a straightforward reversal of the causal order. It nonetheless became orthodoxy, and coaching conversations still carry the assumption long after the strongest sides stopped sharing the profile.

A more honest use

The number is genuinely useful as a description of how a match unfolded and as a signal that one side had a particular intention. It is close to useless as a verdict on which side played better, because it is silent on the dimension that decides matches. Reporting it alongside where the ball was held and how long sequences lasted would cost very little and would remove most of the confusion.

The persistence of the bare figure has more to do with broadcast convention than with anyone believing it settles an argument. Treating a convention as a conclusion is the error, not the counting.