The offside line graphic is drawn to persuade you

Those crisp coloured lines look like measurement, and the parts that involve human judgement are exactly the parts you never get to see.

The lines are beautiful. Two of them, laid across the turf in team colours, meeting at a point that decides a goal. Very clean. Very final.

What the graphic does not show you is where the judgement happened. Three separate calls sit behind that picture, and every one of them is a person choosing. Somebody picks the frame at which the ball is deemed to have left the passer's foot, from a feed running at a limited number of frames per second, which means the true moment falls somewhere between two images and one of them has to be chosen. Somebody picks the body part on the attacker that is furthest forward and legal, and somebody picks the equivalent on the defender, which sounds simple until you look at a still of a sprinting full-back with an arm trailing and a knee raised. Then a calibration model turns those pixels into positions on a plane.

Each step has a margin. Stacked, the margin is bigger than the sliver the lines are being asked to adjudicate. And yet the output is rendered with the confident geometry of a surveyor's drawing, in high definition, on a big screen, and the crowd is invited to accept it as measurement.

Semi-automated systems help with speed and consistency, and they genuinely reduce the worst delays. They do not remove the frame-selection problem, and they do not remove the fact that a limb-level margin of a few centimetres is being used to make binary decisions in a sport where the ball is bigger than the margin.

So my view is unpopular in both directions. Keep the technology, because the alternative was a linesman guessing from twenty metres away with a striker moving at full speed. Change what gets displayed. Show the tolerance. Draw a band rather than a line, and if the attacker sits inside the band, the on-field decision stands. That is more or less how a physicist would present it and more or less how cricket already handles a ball clipping the stumps on review.

It would also end a specific and tedious argument, the one where a goal is disallowed for a shoulder and half the ground insists the technology proved it while the other half insists the technology is nonsense. Both would be wrong, which is usually the sign that the picture was too confident.