A quick review of why the VAR offside assessment implementation allows for manipulation in these tight attacking instances (which are very common for both us and Liverpool).
1. The VAR chooses the frame which will be used to determine offside.
It should be assessed at the point of contact by the passing player, but here it appears the VAR official chose to select a later frame where the ball is well off the passing player’s boot (footballs compress on contact and expand after contact; the expansion is magnified by the frame rate of the camera being used, elongating the image of the ball well beyond it’s actually shape in real life). A VAR official should have been trained to understand these videography related distortions.
2. The VAR places the vertical line on the frame that represents the attacker index position to be used by the camera distortion correction algorithm to project the 2D marker line on a 2D representation of a 3D space.
It is very easy to place the vertical index line slightly ahead or slightly behind the part of the attacking player’s body (that they can legally score with) closest to goal due to the resolution of video and having to interpret 3D space in a 2D image.
3. The VAR places the vertical line on the frame that represents the last non-goalkeeper defender index position to be used by the camera distortion correction algorithm to project the 2D marker line on a 2D representation of a 3D space.
Again, it is very easy to place the vertical index line slightly ahead or slightly behind the part of the defending player’s body (that they can legally defend with) closest to goal due to the resolution of video and having to interpret 3D space in a 2D image.
In addition, the nature of capturing complex movement in a 3D space with equipment designed to record 2D imagery means that the frames themselves can distort the relative distance of the players on the field and/or the point of contact with the ball (i.e. there may be no frame that actually captures the real point of contact, so the next best available must be used; there may be no frame that properly represents the positions of players without lense and timing distortion).
There is a margin of error in this system — every system like this has one. But the PL refuses to confirm what that margin of error is, nor do they even acknowledge it is factored in to the assessment.
All of these individual determinations mean there is significant variance between VAR offside assessments from incident-to-incident, and means there is broad leeway for manipulation of the outcomes of these assessments. There are ways to improve the accuracy of their assessments, especially in the context of the spirt of the game (favouring goal scoring, rather than defending) — such as using a wider line for the defender in proportion to the margin of error of the system (if the thinner attacker marker line is within or behind that line, the attacker is deemed onside) and/or simply using only feet to determine the markers for attacker and defender (as normally no great advantage is realise by leaning slightly forward with your shoulder in these situations, as was this case with this instance), but the PGMOL and PL seemingly refuse to entertain making those changes.