Packing Rate

Definition

Packing Rate measures the number of opposing players (defenders) that a pass or dribble bypasses. A pass that "packs" 3 defenders means 3 opponents were between the ball's starting position and its destination, and are now behind the ball after the action.

History & Origins

Packing was developed by Stefan Reinartz (former Bayer Leverkusen and Germany midfielder) and the data company Impect GmbH (founded in Cologne, Germany) around 2016. Reinartz transitioned from playing to analytics and co-founded the metric as a way to capture what he felt existing statistics missed: the ability of a pass or dribble to eliminate defenders from the play.

The concept gained significant public attention when Impect presented it at football analytics conferences and it was featured in German football media. DFB (German Football Association) reportedly used packing data in their analysis, and the Bundesliga adopted it in broadcast graphics.

Packing was positioned as an alternative to traditional passing statistics (pass completion %, key passes) that Reinartz and Impect argued were misleading — a high pass completion rate might just mean safe sideways passes, while a low completion rate might belong to a player attempting difficult line-breaking balls.

How It Works

At the moment of a pass or dribble, identify the positions of all opposing players
Draw a line from ball origin to ball destination
Count how many opposing players were between the ball's start and end positions (specifically, players who are now "out of the play" — behind the new ball position relative to their own goal)
That count is the packing value for that action

Variants

Packing: defenders bypassed by any action (pass or carry)
Impect: a specific subset — defenders bypassed when the ball reaches the final third or the penalty area, weighted by the importance of the bypassed defenders (bypassing a center-back is worth more than bypassing a winger)

What It Reveals

Line-breaking ability: identifies players who consistently play passes that cut through defensive lines, even if those passes aren't assists
Dribbling effectiveness: a dribble past 3 opponents has a packing value of 3, quantifying the impact of take-ons
Midfield penetration: particularly useful for evaluating deep-lying playmakers and progressive midfielders who don't show up in goals or assists but break down defenses

Limitations & Debates

Proprietary: unlike xG or xT, Impect's packing data is not publicly available and the exact methodology has not been fully open-sourced, limiting independent verification and adoption
No connection to goal probability: packing counts defenders bypassed but doesn't directly connect to scoring likelihood — bypassing 3 defenders on the wing isn't necessarily more valuable than bypassing 1 defender centrally near the box
Superseded by xT and VAEP: the analytics community largely moved toward value-based frameworks (see xT - Expected Threat and VAEP - Valuing Actions) that connect ball progression directly to goal probability, which packing doesn't do
Requires tracking or detailed positional data: computing packing needs knowledge of all player positions, not just the ball, making it harder to derive from standard event data
Doesn't account for what happens next: a line-breaking pass into a dead end (no shooting angle, no onward options) gets the same packing value as one that creates a clear chance

Relationship to Other Metrics

xT (see xT - Expected Threat) → values ball progression by goal probability, not defenders bypassed. xT has become the more widely adopted alternative
Progressive passes (see Progressive Passes) → a simpler threshold-based approach to identifying forward passes, without counting defenders
VAEP (see VAEP - Valuing Actions) → values all actions including line-breaking passes, using ML rather than geometric counting

Key People

Stefan Reinartz — co-creator, former professional footballer turned analyst
Impect GmbH — the company behind packing, based in Cologne

Resources

Impect GmbH — official provider of packing data
Various conference presentations by Reinartz and Impect at OptaPro Forum and other events

Tags: #football #analytics #packing #metrics #passing #ball-progression

History & Origins

How It Works

At the moment of a pass or dribble, identify the positions of all opposing players

Draw a line from ball origin to ball destination

Count how many opposing players were between the ball's start and end positions (specifically, players who are now "out of the play" — behind the new ball position relative to their own goal)

That count is the packing value for that action

Variants

Packing: defenders bypassed by any action (pass or carry)

Impect: a specific subset — defenders bypassed when the ball reaches the final third or the penalty area, weighted by the importance of the bypassed defenders (bypassing a center-back is worth more than bypassing a winger)

What It Reveals

Line-breaking ability: identifies players who consistently play passes that cut through defensive lines, even if those passes aren't assists

Dribbling effectiveness: a dribble past 3 opponents has a packing value of 3, quantifying the impact of take-ons

Midfield penetration: particularly useful for evaluating deep-lying playmakers and progressive midfielders who don't show up in goals or assists but break down defenses

Limitations & Debates

Proprietary: unlike xG or xT, Impect's packing data is not publicly available and the exact methodology has not been fully open-sourced, limiting independent verification and adoption

No connection to goal probability: packing counts defenders bypassed but doesn't directly connect to scoring likelihood — bypassing 3 defenders on the wing isn't necessarily more valuable than bypassing 1 defender centrally near the box

Superseded by xT and VAEP: the analytics community largely moved toward value-based frameworks (see xT - Expected Threat and VAEP - Valuing Actions) that connect ball progression directly to goal probability, which packing doesn't do

Requires tracking or detailed positional data: computing packing needs knowledge of all player positions, not just the ball, making it harder to derive from standard event data

Doesn't account for what happens next: a line-breaking pass into a dead end (no shooting angle, no onward options) gets the same packing value as one that creates a clear chance

Relationship to Other Metrics

xT (see xT - Expected Threat) → values ball progression by goal probability, not defenders bypassed. xT has become the more widely adopted alternative

Progressive passes (see Progressive Passes) → a simpler threshold-based approach to identifying forward passes, without counting defenders

VAEP (see VAEP - Valuing Actions) → values all actions including line-breaking passes, using ML rather than geometric counting