Home/Soccer Analytics/Passing Networks with xG Contribution Overlay

Passing Networks with xG Contribution Overlay

Passing MetricsLevel 3 — Advanced

What It Is

A passing network shows the volume and direction of passes between players, but standard networks don't capture value. By coloring each passing connection (or node) based on the xG generated from possessions that passed through that connection, you can identify which passing links actually led to dangerous attacks — not just which players passed to each other most frequently. High-volume connections that generate no xG are structurally busy but tactically sterile.

Correct Execution

Construction: for each pass event, tag the possession sequence it belongs to; compute the xG of the possession's final shot (or 0 if no shot); assign that xG back to every pass event in that possession chain. Aggregate by passer-receiver pair to get "xG contributed per passing connection." Overlay on the network as color intensity or edge thickness. Nodes (players) can also be sized or colored by their total xG-contribution across all possessions they participated in.

Progression Levels

Diagnostic Tree

Coaching Cues

  • "The red connections on this network are the ones that matter. The gray ones are just ball-keeping." — Ted Knutson, 2018
  • "Volume is noise. Value is signal. Color by xG."

Common Errors

  1. Using shot count instead of xG to color connections: Shot count rewards volume; xG rewards quality. A team that takes 15 low-quality shots looks different from one that takes 5 high-quality shots.
  2. Assigning xG only to the final pass in a sequence: Every player in the possession chain contributed — distribute xG credit across the chain.

Sources

  • Ted Knutson, Barcelona Coach Analytics Summit, YouTube, 2018-11-18 — described xG-overlaid passing networks for half-to-half tactical analysis; used Arsenal as example (slow first half → high-xG-connection second half)