The Premier League’s official statement said VAR missed an extremely offside Enzo Fernández in last month’s Manchester derby because officials were focused on Erling Haaland. Broadcast replays showed Haaland’s movement so closely observed that referees conceded their attention contributed to the oversight.
During the play, Josko Gvardiol moved up the opposite side of the box while Haaland adjusted his strides to remain level with the ball — a half-step with his left, then a lunge with his right — arriving where required when the cross came. When Gvardiol curled the ball around Patrick Dorgu from roughly 30 yards, Haaland was, by modern measurement, three centimeters behind the line of Dorgu’s toe.
Strikers with a talent for timing runs are often overlooked. The piece recalls AC Milan forward Filippo Inzaghi, whose offside-hunting prompted Sir Alex Ferguson to joke the player “must have been born offside,” a remark Inzaghi answered by saying he was “born a meter onside.”
Measuring that particular skill is complicated. Counting offside flags alone is misleading: the analysis notes Haaland has been caught offside 25 times in the Premier League, the same total as Crystal Palace’s Ismaïla Sarr. Ratios such as flags per goal also falter — Chelsea and Arsenal midfielder Jorginho scored 21 goals with zero offside calls — and raw comparisons fail to account for differing opportunities among players like Nottingham Forest wing back Daniel Muñoz, Jean-Philippe Mateta or Yoane Wissa.
The article proposes a different approach: an on-ball expected-offside model dubbed xOffside. Such a model evaluates each pass in context and finds patterns familiar to observers: longer, faster passes are likelier to draw flags; chipped or lofted balls yield more offside calls than comparable ground passes; and the pass angle matters, with straight-ahead balls more vulnerable to traps than diagonal ones. Because ball location alone is insufficient, the model incorporates phase-of-play data to improve its predictions.
