• Welcome to OGBoards 10.0, keep in mind that we will be making LOTS of changes to smooth out the experience here and make it as close as possible functionally to the old software, but feel free to drop suggestions or requests in the Tech Support subforum!

Totally Unofficial 2021/22 Premier League Thread (NWT)

yeah, I was just curious as to Rashford's two saved shots before I started watching. Saw some twitter talk and match reports saying he is a wasteful striker but then I see an xG of 1.2 and 1 goal and am wondering how big of misses were the chances in question?

And what was his xG on his goal? I would have thought that xG would have been pretty high as he was through, not really under pressure, and 1v1 against the keeper from a decent angle.

a goal from 3 shots on target isn't a bad return at all, without getting into individual xG per shot stats

it's exactly where he is for the season on g/SoT, .33

and considering on season he's got 8 goals with a 7.1 xG, it'd be harsh to criticize him over the term
 
brb, i gotta go argue with different strangers on the internet.
 
FWIW, dv7, Auba's miss yesterday from about 6 yards out into a mostly open net was only an xG of 0.5 according to Twitter stats people
 
FWIW, dv7, Auba's miss yesterday from about 6 yards out into a mostly open net was only an xG of 0.5 according to Twitter stats people
Example 129,098,023 of why xG seems to be an inherently flawed statistic, imo

No way that THIS type of chance is only scored HALF the time. 9/10 at worst, even for an adult league team in Winston Salem.

c1cc2a55fe3545a9c5bfc33a6846cc2d.jpg
 
I don't recall if all of that context is taken into account, such as presence and positioning of defenders etc.
 
I don't recall if all of that context is taken into account, such as presence and positioning of defenders etc.
I read that it did on fbref which uses statsbomb's xG metric

All penalties are .76 xG (regardless of weather, which again makes it flawed)


And I'm not blaming xG but people shouldn't take it as gospel when it clearly has its troubling shortcomings. It does provide a nice general idea of how many chances a side creates tho...but so does SoT....and there is no bias in straight yes/no stats like that one
 
There’s no bias in shots on target but a slow dribbler the keeper easily saves counts as a SoT and a rocket that completely beats the keeper but comes off the post doesn’t.

Any stat by itself is imperfect.
 
other than the "goals scored" stat, of course

but you're also talking to the guy that HATES it when commentators say such a rocket shot off the post was "unlucky"

no it wasn't. the shot just wasn't good enough. that's just lazy commentary and it GRINDS MY GEARS.
 

but you cannot say with a straight face that any metric that says the above chance is only a 50/50 proposition is a true one.

goalkeeper flailing already
the entire goal at his mercy
ball on the deck with no bobble


if that is only a "50/50" chance then xG needs to rework its algorithm. by a lot.



ill start to use xG more but i'll do so knowing it is deeply flawed. it is one of the most flawed "advanced metric" statistics i've ever seen. QBR might even be better...


this post is not an endorsement of QBR.
 
It’s only trying to describe something otherwise not captured in counting stats. No more and no less. Savvy data consumers won’t rely on it and they won’t ignore it. Anybody can go after the top goal scorer in a division, but you can get value at the margins scouting players who have a high xG that haven’t found the net as often as the superstars.
 
It’s only trying to describe something otherwise not captured in counting stats. No more and no less. Savvy data consumers won’t rely on it and they won’t ignore it. Anybody can go after the top goal scorer in a division, but you can get value at the margins scouting players who have a high xG that haven’t found the net as often as the superstars.

Why?

Wouldn't that logically just mean they are "below average" finishers? I mean if they were expected to score, let's make some numbers up: 10.8 goals over the season but only scored 7. Seems like a player that wastes chances to me. No?


I could see xA as a MUCH better metric to bring in an undervalued playmaker, however. But why would I want a player that under-performs his/her xG in actual goals?
 
Lastly before I get back to FIFAing... what does xG provide that the following actual stat metrics do not:

Goals
Shots
Shot on Target
% Shots on Target
Goals per shot
% goals of shots on target
% goals of all shots


I would literally murder a hobo to have a conversation with the developer of xG to figure out her/his method. I think that person is on the cusp on something really great, but it needs plenty of tweaks in its current form to actually resemble anything truly useful in football.
 
Theoretically a player underperforming expectations isn’t necessarily a poor finisher. They could have bad luck. A goalie could make a huge save above expectations, for example. I’ll try and find recent examples, but I believe Jota was one that Pool scouted based on underperforming expectations from early in his career. No club would solely look at it as a single stat, of course, when evaluating a player, but I'd be surprised if talent evaluators looked at players vastly over performing xG and thought “this is a truly elite finisher” instead of “this player is due to regress in counting stats”
 
Theoretically a player underperforming expectations isn’t necessarily a poor finisher. They could have bad luck. A goalie could make a huge save above expectations, for example. I’ll try and find recent examples, but I believe Jota was one that Pool scouted based on underperforming expectations from early in his career. No club would solely look at it as a single stat, of course, when evaluating a player, but I'd be surprised if talent evaluators looked at players vastly over performing xG and thought “this is a truly elite finisher” instead of “this player is due to regress in counting stats”

I wanted MUFC to sign Jota from Wolves. I wanted MUFC to sign Mane from Soton. I wanted MUFC to sign Firmino from Hoffenheim.


But Jota...

2018/19: xG 8.1, Goals 9
2019/20: xG 9.6, Goals 7

Total: xG 17.7, Goals 16


This is not a good example of a diamond in the rough.
 
I give full credit to LFC's recruiting team as they have been truly CRUSHING IT for a few years now.


But I would have expected, xA played big into their decision making, but I only see xG mentioned as in the 2nd quote, basically discounting xG:

As revealed in the New York Times in May, the work done by those behind the scenes at Anfield was key in getting Salah to Merseyside. “Graham’s data suggested that Salah would pair especially well with Roberto Firmino, another of Liverpool’s strikers, who creates more expected goals from his passes than nearly anyone else in his position," it read.

“Missing out on potential stars is a lot less harmful than mistakenly signing players that aren't as good as you thought they would be," he said. "A risk-averse approach with this in mind would look to avoid signing any players with potential red flags, be that in their performance data (for example, a striker massively over-achieving his xG) or in other areas (their injury record or what scouts think of them etc).

https://www.liverpoolecho.co.uk/spo...rpool-transfer-recruitment-fsg-klopp-17369211


And I found this quote near the end of the article a perfect foil of the absolute bullshit known as YNWA:

While FSG did all they could piecing this recruitment team together, without Klopp and his coaching staff spearheading the operation and being the link between analysts, scouts and players, this process would be flawed as it was when Brendan Rodgers was in charge.

How nice.
 
model of buying cheaply from the Championship (something West Ham have done brilliantly this past season).
Not just this season. Cresswell, Fredericks, and Antonio are (or were) regular starters for us and were scouted directly from the Championship.
 
Let me tell you. It’s REALLY (eye roll) fun to watch Taki light it up for Southampton.
 
I give full credit to LFC's recruiting team as they have been truly CRUSHING IT for a few years now.


But I would have expected, xA played big into their decision making, but I only see xG mentioned as in the 2nd quote, basically discounting xG:





https://www.liverpoolecho.co.uk/spo...rpool-transfer-recruitment-fsg-klopp-17369211


And I found this quote near the end of the article a perfect foil of the absolute bullshit known as YNWA:



How nice.

I don’t read that quote as discounting xG. You want to avoid signing players massively outperforming xG, that’s useful.
 
Back
Top