Had an idea regarding VAR whilst watching it not be able decide if a ball had gone out of play, far too close to call; but player is offside by a toenail, 100% certain, so I suppose I'm wasting my time thinking about fair implementation and how to get VAR working £right". It's working just fine. Not a headbutt either, couldn't tell, looked like a dive to me.
Amyway,
The argument against VAR and tight offsides is the speed of player breaking defensive line vs frame speed, accuracy concerns. Frame speed, so the answer is...
Focus on the player who plays the ball and measure how many frames are we guessing from when we guess at when the "ball was played"? (using the Andy Grey analysis of VAR for Sterling's 3rd (?) against example)
1. When was contact made? Ping! Red line drawn.
2. When was contact terminated. Ping Yellow line.
Maybe have a 50 frame max set for this guess.
Then change focus and look at player who is possibly offside and decide, using these pre red, red, and then yellow coloured frames, how many of them is the player, clearly and obviously offside.
Bigger % onside than offside, goal!
Celebrated after lengthy statistical analysis.
All zoom functions allowed and final split of % shown on big screen for crowd to "cheer" or "boo" when it's a near 50-50 split, when permitted under government regulations. Imagine watching those ad breaks when long decisions are being made (video loop of incident shown in lower right corner of the screen for "authenticity" and "integrity")
OR
NO zoom, or multiple replays (30 seconds max), no lines.
If it looks "silly close" give the benefit back to the forward. Go with the line "he can hardly be deemed offside for a toe, can he?" rather than "his toe is clearly and obviously offside and yes, he's obviously seeking to gain an advantage by that" line.
But this way doesn't take into account any calculated tolerance, and we all know how subjectivity is loved by the makers and appliers of the laws of the game/money making scam.
Can of worms whichever way you look at it I reckon.