10 - Ansu Fati

[ajnotkeith]

Sometimes xG is skewed from chasing the game and intent for chance creation.

And then of course counter goals.

Well Newcastle 19 shots and 1.4 xG, they only scored 1 big chance. Low percentage shots.

That GK for Brighton probably let them down.

 

[jamrock]

Did you watch the game?

maybe villa was just way better in this game?

 

[Porque]

Tbf I didn't watch the game.

Just sometimes I feel that the xG data in isolation has gaps.

Im sure the clubs have their own algorithms for this.

 

[Birdy]

Tbf I didn't watch the game.

Just sometimes I feel that the xG data in isolation has gaps.

Im sure the clubs have their own algorithms for this.

It's not scewed.
Look at the value of each of the chances Brighton created. They could have another 2 goals for sure.
Now look at the value of the goals of Villa. There is some crazy overperformance there

The xT shows that both teams were equally dangerous.
A draw would have been a fair result based on both metrics

 

[mc_lovin]

Haha, shows how shit this metric is.

 

[Messigician]

It's not scewed.
Look at the value of each of the chances Brighton created. They could have another 2 goals for sure.
Now look at the value of the goals of Villa. There is some crazy overperformance there

The xT shows that both teams were equally dangerous.
A draw would have been a fair result based on both metrics

Good analysis haters will still call it optabirdy though @KingLeo10

 

[ajnotkeith]

Did you watch the game?

maybe villa was just way better in this game?

I did. I dont get what thats meant to mean though because xG is an objective stat and usually quite accurate re chances created.
3-4 of Villa's goals should never go in. Just crazy low percentage finishes or deflections.

 

[KingLeo10]

Haha, shows how shit this metric is.

It's not a coincidence that the ones who put a lot of stock into xG metrics are tangential/pseudo-statisticians/ pseudo-intellectuals with a chip on their shoulders.

Anyone with any formal training on the matter will recognize that while there is some (to be very generous) value in these analytics, the models for the large part are simply a poor representation of reality. Mostly because: 1) sports isn't that lucrative an enterprise compared to finance, health care etc.; 2) there are constructs such as team pedigree, mental toughness and player's ability to respond/wilt under pressure that simply can't be captured just yet.

 

[jamrock]

I did. I dont get what thats meant to mean though because xG is an objective stat and usually quite accurate re chances created.
3-4 of Villa's goals should never go in. Just crazy low percentage finishes or deflections.

Okay, was just checking if you wanted the catch, to match your stats.

Statistics in isolation no matter how good, never tells the full story.

 

[ajnotkeith]

It's not a coincidence that the ones who put a lot of stock into xG metrics are tangential/pseudo-statisticians/ pseudo-intellectuals with a chip on their shoulders.

Anyone with any formal training on the matter will recognize that while there is some (to be very generous) value in these analytics, the models for the large part are simply a poor representation of reality.

They are well applied and relevant in this case as evidenced by goals scored which are absurd finishes and deflected goals which pretty much never go in.

Brighton with a good GK concede 1 or 2 goals.

 

[ajnotkeith]

Okay, was just checking if you wanted the catch, to match your stats.

Statistics in isolation no matter how good, never tells the full story.

Agree that they need to be contextualised by watching games or they're not useful.

 

[Birdy]

Okay, was just checking if you wanted the catch, to match your stats.

Statistics in isolation no matter how good, never tells the full story.

There are not 'in isolation', as there tons of sophisticated stats you can use to analyze a game apart from xG.
xG is there to measure the creation of each team.
You can talk a lot about how and why such creation happened or didn't happen using tons of other stats

 

[KingLeo10]

They are well applied and relevant in this case as evidenced by goals scored which are absurd finishes and deflected goals which pretty much never go in.

Brighton with a good GK concede 1 or 2 goals.

The way I see it, xG, over a large sample size, is a decent metric for low leverage games, primarily league and domestic cup games.

I have a strong suspicion that once you get into a high leverage KO format such as CL, WC, Euros etc. or big games such as El clasico, the actual results are poorly predicted by xG. You only have to look at how WC 2022 panned out, with xG darlings Germany and Spain not doing shit

So, if it's not a good predictor of the more important games, what value are you really gaining from it?

 

[ajnotkeith]

The way I see it, xG, over a large sample size, is a decent metric for low leverage games, primarily league and domestic cup games.

I have a strong suspicion that once you get into a high leverage KO format such as CL, WC, Euros etc. or big games such as El clasico, the actual results are poorly predicted by xG. You only have to look at how WC 2022 panned out, with xG darlings Germany and Spain not doing shit

So, if it's not a good predictor of the more important games, what value are you really gaining from it?

Cup games affected by unique factors like pressure, different tactical setups, but the main one would be the quality of players is much different to the usual.

If xG is an absolute average of average scoring rate it stands to reason that much better players are going to overperform it very regularly. However, this becomes even more skewed when its world class vs world class.
xG numbers weren't modelled off such high level games, they were modelled off average scoring rates, so it will be harder to apply them to a much higher level than they were formed off.

 

[fergus90]

We hammered Brighton. Always a guaranteed 3 points when they visit Villa Park.

Glad they made the score line a bit closer on xG though.