It's all about averages over a period of time. The most you can take something from, for example:
Man City 2.78 vs. Tottenham 1.13 <---------------- and perhaps the game ends 2-2
From this all you can really infer is that City were the better team at creating high quality chances and deserved to win.
2.78 indicates City probably scored 2-3 goals, but within 90 minutes of football the variation could be wild, maybe no goals or 5 goals.
If your time period extends into thousands of minutes, then the variation looks much tighter.
One example that uses these stats has that with only 12 games in, that City should have expected to score 37.58 goals this season and we have scored 35 goals.
100*35/37.58 is about 93% accurate. That's pretty good for a stat that measures long term data.
It's even closer with how many we should have concede, stats say 13.48 when we have conceded 13, about 96.5% accuracy.
One could also possibly infer within the game how good our finishing was, maybe not completely, but to some extent.
If I see an xG for city that says 3 or higher, and we score 1 or no goals, then without even watching the game I might assume we finished poorly.
Man City 2.78 vs. Tottenham 1.13 <---------------- and perhaps the game ends 2-2
From this all you can really infer is that City were the better team at creating high quality chances and deserved to win.
2.78 indicates City probably scored 2-3 goals, but within 90 minutes of football the variation could be wild, maybe no goals or 5 goals.
If your time period extends into thousands of minutes, then the variation looks much tighter.
One example that uses these stats has that with only 12 games in, that City should have expected to score 37.58 goals this season and we have scored 35 goals.
100*35/37.58 is about 93% accurate. That's pretty good for a stat that measures long term data.
It's even closer with how many we should have concede, stats say 13.48 when we have conceded 13, about 96.5% accuracy.
One could also possibly infer within the game how good our finishing was, maybe not completely, but to some extent.
If I see an xG for city that says 3 or higher, and we score 1 or no goals, then without even watching the game I might assume we finished poorly.