Pass Networks
A pass network plots each player's average on-pitch position as a node, with node size and connecting line thickness showing how much they passed and to whom.
Why it matters. It reveals a team's shape and structure on the ball, showing which players the team builds through rather than just who completed the most passes.
Figures from xG Stat’s own models, built on Wyscout event data. Updated .
How the network gets built
Every player who featured gets a node, positioned at the average location of their touches on the pitch. Two properties layer on top of that position.
- Node size tracks pass volume. A player heavily involved in possession draws a larger node than one who saw the ball rarely.
- Line thickness tracks combination frequency. A line between two players thickens with every pass exchanged between them, in either direction.
- A toggle switches views. Average position shows just the dots, for reading overall shape. Network view adds the lines, for reading who combined with whom.
- A minimum-passes slider declutters busy matches, hiding faint, one-off connections so only the frequent, structural links remain visible.
The pitch background never changes. Only the dots and lines drawn on top of it do, which is why comparing two matches side by side is mostly a matter of comparing shapes.
Start with the average position view before switching on the lines. A team's overall shape is easier to judge from the dots alone, and the network view layered on top is easiest to read once that base shape is already familiar.
Three shapes worth recognising
A compact, tightly clustered network. Every node sits close together in a small area of the pitch, with short, thick lines connecting them. That shape describes a team that kept the ball in a confined zone, often through central midfield, rather than using the pitch's width or depth.
A stretched, wide network. Nodes spread from touchline to touchline with long connecting lines running between them. That describes a team building through width, moving the ball from flank to flank rather than funnelling it centrally.
One central node with unusually thick lines to almost everyone. A single player sits in the middle of the network with heavy connections running out in every direction, thicker than any other player's links. That is a deep playmaker the team funnels possession through, whether the ball is going forward, sideways, or back.
What structure shows that completion rate cannot
A pass completion percentage is a single number describing accuracy. It says nothing about where those passes went or who the team relied on to make them.
Shape describes a system, not a stat line. Two teams can complete passes at the same rate while one builds through a tight central triangle and the other spreads it across back four and both wings. The completion percentage cannot tell them apart. The network does it immediately.
A missing link describes a plan, not an accident. If two players who would normally combine show a thin or absent line, that gap is visible on the network in a way no individual pass count would surface. It points directly at a tactical instruction or a matchup problem worth checking against the match itself.
A stretched network describes reach, not effectiveness. Width shows where the ball travelled, not whether it did anything useful once it got there. Read a wide network alongside chance creation figures rather than treating spread alone as proof of a good attacking performance.
Two misreadings to drop
"Average position means the player stood still there." It is a mathematical average across every touch in the match, not a location the player occupied. A box-to-box midfielder who covers the full length of the pitch can still average out to one dot near the centre circle, despite never actually standing there for long.
"A thick line between two full backs proves good chemistry." It can just as easily mean the team builds down that side by habit. That is a set instruction, not a personal understanding between the two players. Check the pattern across several matches before reading a thick line as anything more than a repeated route the ball took.
One team's passing network
2025/26
Manchester United 2-0 Manchester City, 17 January 2026: Manchester United's network. Each node sits at a player's average position, sized by how much they passed, and each line is thicker the more often that pair combined.
FAQ
Related terms
See Pass Net across the Premier League
Every player, team and match measured on the chances created rather than the scoreline.