Newcastle Passing Networks

How to read this chart

• Each circle represents a player, positioned at their average passing location on the pitch.
• Circle size reflects the number of passes attempted -- bigger circles = more passes.
• Lines between players show passing connections. Thicker lines = more passes between those two players.
• Line colour indicates expected threat (xT) gained: green = moves the ball to more dangerous zones, red = backwards/sideways.
• Only starting XI and connections of 3+ passes are shown to reduce clutter.

Methodology

• Pass event data sourced from WhoScored (Opta events). Coordinates normalised to a 100x100 grid.
• Pass receivers inferred from the next touch event by the same team (standard analytics approach).
• Expected threat (xT) values from Karun Singh's xT model, using the Laurie Shaw EPV grid.
• xT gained per pass = destination zone value minus origin zone value. Positive = progressive.

Season aggregate controls

• Size by -- what drives each player circle's radius. Passes shows total volume, Progressive shows successful passes that gained positive expected threat (i.e. moved the ball into more dangerous areas), xT shows total expected threat generated, and Completion shows pass accuracy as a percentage.
• Players -- how many players to display. Top 11 shows the most-selected XI, ranked by appearances with passes as a tiebreaker; a spatial deduplication step prevents two players who occupy the same position (e.g. rotation left-backs) from both appearing. Regulars (15) and Extended (20) show the wider squad ranked by appearances with no spatial filtering.
• Position -- filters players by their registered position from SportMonks data (e.g. Centre-back, Winger, Centre-forward). Falls back to average passing position on the pitch for unmatched players.
• Min. apps -- filters out players below an appearance threshold. Useful for removing cameo appearances that distort the network.
• Game state -- filters by scoreline. "All" includes every minute. "0-0" shows the shape before any goals (the manager's intended plan). "Winning" and "Losing" show how shape changes when ahead or behind. States with insufficient data are greyed out.
• Values -- toggles between raw totals and per-90-minute normalised values. Per 90 adjusts passes and xT by actual minutes played (including stoppage time), making volume-dependent stats comparable across players with different playing time.