I don’t know how that’s the takeaway you’ve pulled from that. xG stats suggested they were overperforming last year and they would be unlikely to do as well this season. So far, that appears to be correct. I’m not a lover of expected ‘x’ stats but the models are working as intended in this instance.

Exactly, he’s a good young player with potential but he needed consistent minutes to continue developing and he was never going to get that at City while we’re competing for everything. We’d have preferred a buyback or to loan him if possible but equally you’re not going to say no to £40M+ for a fairly unproven youth player.

They’re not asking if VAR allows the attacker any leeway in offside calls. They’re asking about the accuracy/precision of the technology in generating these images, the answer to which is that the margin of error hasn’t been publicly disclosed.