PPDA rewards teams for pressing badly

Passes allowed per defensive action drops when a press fails as easily as when it succeeds, which makes it awkward to read.

The metric counts opposition passes in a defined area and divides by the defensive actions you attempt there. Fewer passes per action means a more aggressive press. It is easy to calculate, it is available everywhere, and it is used as though it measured how good a press is rather than how often one was attempted.

Those are not the same thing and the gap is where the trouble starts. A team that presses in a coordinated way and wins the ball high records defensive actions. A team that presses chaotically, with players flying in and getting bypassed, also records defensive actions, arguably more of them, because a failed challenge is still a challenge. Both look similar in the number.

Worse, an excellent press can push the number in the wrong direction. If a side presses so well that opponents stop trying to play out and just kick it long, then the passes in that zone vanish, the defensive actions vanish with them, and the ratio behaves strangely. You have coached the opposition out of the phase the metric is measuring and the metric cannot see it.

There is also the game-state problem, which affects nearly every possession-adjusted statistic and is rarely handled properly. Teams that lead press less. Teams that trail press more. So the number partly measures scoreline history, and comparing a side that leads often with a side that chases often tells you something about their seasons rather than their coaching.

I still look at it, to be clear. Not as a rating but as a flag. A large change in a team's figure between two months is worth investigating, because it usually means something concrete happened: an injury to the forward who leads the press, a change of shape, a manager. The level tells you little. The movement tells you where to look.

The measurement I would prefer does not exist publicly, and it is the share of opponent build-up sequences that fail to cross the halfway line. That is closer to what a press is trying to do, and it does not care whether you achieved it by tackling or by making the pass impossible.

Until then we are counting attempts and calling it quality.