Unai Emery

[paul_e]

As I've said many times, it is a load of bollocks because IT'S NOT WHAT THE STAT IS MEANT FOR.

xG was designed to be an improvement on just having the number of shots/shots on target to show how well you've played in attack, that's it. Over long periods there is often a degree of of correlation between xG and how many goals are scored, much like there is a correlation between shots and goals but tables like that are utter nonsense.

Specifically to the Rogers chance mentioned, all xG uses for it's calculation is distance from goal, angle, ball height and shot type. So Rogers had a shot from fairly central, about 4 yards out with his left foot and the ball was very low to the ground, with no more context you'd expect that chance to be scored most of the time.

[Garyth]

Can anyone explain to me that that’s not a load of bollocks?

I agree, It’s definitely horse shit.

Lads, it's not that hard (and been explained by many people).

XG doesn't provide a value for 'Is your team any good', it only gives you 'from thousands of similar shots, how many have gone in' - no adjustment for game state, or specific player involved, etc etc.

Stop pretending it's some sort of judgement saying Villa are bad at football - in fact, it's saying the opposite .... eg. Rogers is outperforming what is usually expected in the positions he gets into. The top scoring players in the league each year usually outperform their xG over the season - in other words, they are (surprise!) better than the average player at scoring. Similar with the teams that win it.


One (of many) explainer: https://learning.coachesvoice.com/cv/expected-goals-xg-explained/