Man City lost 3-1 to Lyon in their Champions League quarterfinal on 15 August 2020, but the expected goals metric paints a different story of the match-City dominated in terms of possession and shots, and actually finished with a xG of 2.68, as opposed to OL’s 0.87 xG.
Stats- xG score of individual goals in City-OL
0-1 Maxwell Cornet 24 (xG 0.11)
1-1 Kevin de Bruyne 69 (xG 0.06)
1-2 Moussa Dembele 79 (xG 0.11)
1-3 Moussa Dembele 87 (xG 0.51)
xG of Raheem Sterling terrible 86th minute miss: 0.59
Other notable City chances in terms of xG: Jesus 77th minute (0.53) and Sterling 45+1 (xG 0.50)
Edit: Here is a second xG model for the game courtesy SB, and they have City with a xG of 2.42, versus 1.07 for Lyon. It also has the Sterling miss with a xG of 0.87, and the first Moussa Dembele goal at a xG of 0.37, potentially taking into account the positioning (or lack of it!) of the respective keepers.
One would think the xG for the Sterling miss ought to be a little higher. Penalties tend to have a xG around 0.75. This is right up there in terms of a chance, perhaps even close to 0.8
I understand xG models can vary, but think this doesn’t take into account the fact that there was no keeper present to make the save for the Sterling shot. It’s the only way I think the miss would have a xG of 0.59 and not a 0.8.