It may be the best tool we have and it may paint an "accurate" picture and it may "indicate" something. But those are qualitative statements. They do not answer the quantitative question of "what numbers do we need to be looking at to make conclusions with an acceptable degree of certainty."
I think that scientifically and philosophically, it is tricky to say the difference between a model's predictions and actual outcomes are the result of a hidden variable (ability to convert chances) without corroborating that hidden variable independently and without judiciously accounting for other sources of error.
It is also an unnecessary argument. Ronaldo's actual goals are as low as they've been in years. His xG is also as low as its been in years. This is happening while he plays for a team that also has a low xG overall. This is all bad and much more robust than "maybe he's not shooting so good."
He is shooting as 'bad' as Diogo Jota and Sadio Mane by this standard but I can't imagine Klopp is too concerned about that since his team is scoring lots of goals and has high 'expected goals.'
We absolutely agree that the xG data is flawed, it is way to many variables to make an 100% accurate prediction for each and every shot taken by using averages, so if we were looking at numbers alone, without watching the games and adding the two together, it would in theory be possible to have extreme exceptions, yes. The bigger the sample size the more accurate the xG will be though. I've watched all our games and I have no reason to believe his chances this season has been much harder than the average over the 20 games he's played in the league, and I think is finishing has looked pretty toothless with a couple of exceptions.
It's too easy to dismiss stats, because deviations are possible and common to a certain degree for smaller sample sizes.
Only stupid people IMO.
Anyone who played the game won’t care a bit about stats in the bigger picture.
Posters talking about xG, when his only clear cut chances came in Watford game, West Ham first game and that one he was 1 on 1 with the gk. The rest of the service and quality of crosses, positions where he took the shot massively skew the data
Interesting discussion technique here. First you generalize everyone with a different opinion than you and give them unflattering qualities regarding their IQ and belittle their footballing experience, then you go directly from letting us all know how it is about the bigger picture to giving us details about the 2 good chances he's missed this season in your opinion.
Well done! I bet you are popular at parties and social settings.
I'm not going to try to prove to you that I'm not stupid, but I like to back up my arguments with stats and data in addition to what I see, so I obviously must be a bit dim.
I have played football for 35 years though, 20 of them as a striker on senior level, so I can say I have kicked a ball a few times.
I think xG correlates pretty good with what I see with my own eyes most of the time over a span of games. You do obviously have cases like the 2 gifs Iker posted earlier here that makes it not 100%, but like mentioned, over a few games it is a good analysing tool in my opinion.
If you think Ronaldo has looked leathal and efficient in front of goal in the league this season, we have very different views on what it means to be leathal and efficient in front of goal.