Why we should use xG, over results, when judging managers

A thread

xG, short for expected goals, has been a metric which has been steadily gaining popularity, and is used to calculate the quality of a shot, with Statsbomb’s model taking into account variables such as distance, body part, the positions of defenders and the keeper and shot height

It can have a variety of uses, such as figuring out which players are getting into good goal scoring positions, which players have been good/lucky or bad/lucky finishers, or how well a team is playing or how much a team has dominated a game.

It is because of that last reason, that I think xG is better to use to evaluate a managerial performance than simply results. I interpret a manager’s job, at least when setting up their team, as trying to maximize the performances of the talent they have at their disposal.

In other words, how can the manager get the most out of their squad? How can they play in such a way that plays to their players’ strengths? How can they over perform the most with the talent they have? How can they add more to this team than is already available in their talent?

So for example, someone like Chris Wilder over performed with his talent last season, meaning he did a terrific job. But why do we need xG? Surely we can just look at a team’s results, comparing it to their talent level, and use that to assess managerial ability?

Because results take into account luck, and more specifically, finishing and goalkeeping. While you can say a team deserved to lose because they didn’t finish or because the opposition goalkeeper played very well, that's not something the manager can control.

Results can be influenced by individual moments of failure/brilliance, moments the manager has no control over. xG does not reflect these individual moments to the same degree has actual results, and that is why it is more important in this discussion.

If a manager sets up their team to dominate, and they do, but don’t get the result they wanted due to poor finishing or good goalkeeping, then the xG is more reflective of the managerial and team performance as a whole, not the result

xG offers a better representation of overall team performance because, which the manager controls, as it’s less influenced by individual moments than results, which is more impacted by events outside the manager’s control, such as finishing and goalkeeping

Part of the reason I made this thread is because of the replies to Gasipo's Brighton thread: https://twitter.com/gasipo_opinions/status/1389592050119630849?s=20, but also for replies to my Pirlo thread (I think Pirlo's been fine) and my thoughts on Mourinho (who I think had been poor at Tottenham)