Just over performance v reversion to the mean is all. Statisticians invoking the Gods.
Alisson over performs v XGA, Liverpool give up very few chances of any quality, let alone high quality... and the worry was always for Dan Kennett that it’s revert to the mean and scupper our chances to some extent.
Nothing more complicated than that really.
I understand, but I’m just not in agreement that Alisson’s goals conceded
has to revert to xG. He may well revert to the mean, but that will be because he’s not quite as good as the current stats show. Then again, Alisson may outperform xG because he’s a better quality shot stopper than the ‘mean’.
Messi outperforms xG by 20% I read somewhere, which is because he is a higher quality player. Morati has a high xG but has been poor at finishing. The variables involved in xG don’t take into account the quality of player/keeper. Also, each chance is independent of the next chance (usually - unless it’s in the same move) so like tossing a coin, if you get a heads that doesn’t make the next toss more likely to be tails.