It's a decomposition, just as any other analysis. It tries to separate the two cases of not scoring and not creating with not scoring and creating. Two meaningfully different scenarios.
Look at the way it's used in relation to Chamberlain. This is my point. His actual output is non-existent, but his theoretical output is world class.
Much of this pseudo-scientific crap is purely descriptive.
I'm already pissed that nobody commented on this important post yet
Don't especvially want to get into a debate about statistics but that's baseless and in many instances simply wrong.
If I'm wrong I'll admit it. I just don't see how almost scoring a goal but losing means you have a better chance of winning in the next match. xG by itself seems a fallacy to me.
There is a demonstrated correlation between xG and goals scored. And the more goals you score, the more likely it is you'll win games.
Yeah it's abused a lot, that I agree with.
That isn't actually what you said. You said " If you almost score a goal, that has no bearing on anything". A high xG correlates with more goals scored. If you're taking lots of shots from dangerous areas and missing then it's likely as long as you keep doing that some of them will start going in eventually. There is a phenomenon called reversion to the mean, unless there's a very obvious reason (i.e you have a complete donkey up front rather than just an average striker) things will usually "turn".
You could have done so much better with 'I X-pect this thread to get locked soon.'
Finally some sense to this thread
United and City made some very good signings which reflects in improved results. We have signed Lukaku (2nd highest goal scorer), Matic to bring stability to the team. City have signed Walker, Mendy who have improved them immensely and Ederson who isn't a clown keeping goals.
It isn't. Both United and City were creating many chances, especially City, and did not concede many chances. The ratio between created and conceded chances was very good which is an excellent basis to build from. Both teams were better than the results suggested. It is the same with Liverpool now. They aren't as bad as the results suggest. Which doesn't mean that they can compete for the title. It means only that they can go on a good run and compete for the top 4.
.Good general explanation, thanks.
I thought he used Chamberlains stats from Arsenal where you'd expect most players to have a high xG due to the type of football they play. Tbf he hardly used xG in that argument either, more just picked the stats to suit his argument about completed dribbles etc which as someone said could be a fullback dribbling 10 yards before turning round and passing it backwards.
I‘m sure you could describe the chick at the bar in metric terms and have a great discussion about your chances, but basically I just want to feck her.
Huh?
I‘m rather a drunk „He‘s shit“ guy
I think it's better for seeing how a team will progress from season to season as the sample size is then big enough. If a top manager uses the data in front of them well enough they should be able to identify the problems.
But the challenge that Klopp faces is that there is no point in reducing the opponent's xG by playing deep and conservative, which effectively nukes his side's xG. Plus, playing deep doesn't prevent Lovren from making a brain-dead mistake.I think it's better for seeing how a team will progress from season to season as the sample size is then big enough. If a top manager uses the data in front of them well enough they should be able to identify the problems.
Take Liverpool this season, they have a high xG themselves, however they also have a high rate of xG against them, meaning they are offering up good chances to the opposition.
That should lead to you wanting to add a more solid base to lower the opponents xG, however they've got Klopp so they're screwed
And that's great. You should watch the game however you deem fit.
I see why you're sad, Chris.
I might be talking out of school here because I don't understand the xG stuff entirely but from what I have read would it not be a good method of determining that a player like Ozil tends to create better chances than someone like Valencia for example.
1. No statistical model is perfect. There will always be outliers that defy common sense. That is not an indictment on the xG model which overall shows good correlation.
Which is why if he wants to keep the same style he should focus his budget on better defenders
I'm sorry.
Quote yourself again. Maybe enough people didn't see it.I'm already pissed that nobody commented on this important post yet
Ooooh. Put those claws away before you scratch something else.
I know it's juvenile but it really tickled me.
Oh, I didn't say it wasn't funny. It totally was, just in a very bitchy way but that is also cool.
Oh, I didn't say it wasn't funny. It totally was, just in a very bitchy way but that is also cool.
Touche!
