Here's the expected save-% for the Liverpool goals using said PSxG:
1st goal (Gakpo): 45%
2nd goal (Nunez): 27%
3rd goal (Gakpo 1): 4%
4th goal (Salah): 73%
5th goal (Nunez): 74%
6th goal (Salah): 33%
7th goal (Firmino): 55%
Here's the City game:
1st goal (Foden): 85%
2nd goal (Haaland): 86%
3rd goal (Haaland): 12%
4th goal (Foden): 12%
5th goal (Haaland): 97%
6th goal (Foden): 21%
Looking at the replay of these goals you can easily just write off post-shot xG as pure non-sense.
Good analysis. Think the stat is arse-over-tit really.
1st goal (Gakpo): 45% - this is too high, it's a good finish, bent round the keeper into the far corner with decent pace from close range.
2nd goal (Nunez): 27% - this one was unsaveable, it's a 1% Banks-esque chance.
3rd goal (Gakpo 1): 4% - more saveable than the first two, but he does everything right and it's just a very clean finish.
4th goal (Salah): 73% - totally unsaveable
5th goal (Nunez): 74% - totally unsaveable
6th goal (Salah): 33% - chaos in front of him
7th goal (Firmino): 55% - most saveable, but he's unlucky.
It's pretty useless here, more just indicates some great finishing and some woeful defending. He has no right to save any of those goals, with only the 7th open for debate.
1st goal (Foden): 85% - no chance, top corner from about 10 yards
2nd goal (Haaland): 86% - no chance, free header from 6 yards
3rd goal (Haaland): 12% - no chance (albeit a very proactive keeper could possibly have intercepted the cross)
4th goal (Foden): 12% - no chance
5th goal (Haaland): 97% - keeper can only get lucky here, arguably he should have been positioned more centrally and it would have hit him. Again it's 1 City player getting a free shot in a box with 7 other United players.
6th goal (Foden): 21% - no chance
Again I can't really fault him.