Expected Goals Against (1 Viewer)

[skybluecam]

Would anyone like to guess where we rank in the expected goals against (xGA) table? For reference, we have currently conceded 34 which puts us 18th in the actual goals against table (where lower = more conceded).

Answer:

Spoiler

5th, with only 23.0 xGA, behind only Leeds, Burnley, Sheff U and Millwall.

 

[TomRad85]

Would anyone like to guess where we rank in the expected goals against (xGA) table? For reference, we have currently conceded 34 which puts us 18th in the actual goals against table (where lower = more conceded).

Answer:

Spoiler

5th, with only 23.0 xGA, behind only Leeds, Burnley, Sheff U and Millwall.

I know we're high, as in good. Which I know suggests unlucky but I don't believe the errors we make which do lead to the goals we concede are unlucky tbh.

 

[Gint11]

5th meaning we shouldn’t be conceding many?
If so then that makes little sense. Just googled how X.GA is measured such as type
of chance, shot location, defense and goalkeeper strength and we are poor at all of
It.

 

[Nuskyblue]

I know we're high, as in good. Which I know suggests unlucky but I don't believe the errors we make which do lead to the goals we concede are unlucky tbh.

Agreed.

At a guess I'd say we're in the top 6 (good) but like you I think the xG doesn't tell the whole story. Our calamitous defending really has fucked us this season to date.

 

[skybluecam]

I know we're high, as in good. Which I know suggests unlucky but I don't believe the errors we make which do lead to the goals we concede are unlucky tbh.

Well having had some truly shocking keeper performances explains a lot of it. And anecdotally we’ve spent a lot of time behind in games where the oppo are likely to sit back more.

But 5th vs 18th is still an insane difference to me.

 

[Evo1883]

Would anyone like to guess where we rank in the expected goals against (xGA) table? For reference, we have currently conceded 34 which puts us 18th in the actual goals against table (where lower = more conceded).

Answer:

Spoiler

5th, with only 23.0 xGA, behind only Leeds, Burnley, Sheff U and Millwall.

Many people stated earlier in the season we were conceding shite goals .


Most of our stats have been in the right place all season as @Frostie has shown several times

 

[Earlsdon_Skyblue1]

They should recalculate that based on what we actually see every week.

 

[stupot07]

xG is still a load of nonense. We battered Plymouth yesterday 4-0, yet our xG was only 1.28.

It's just arbitrary nonsense.

 

[skybluecam]

xG is still a load of nonense. We battered Plymouth yesterday 4-0, yet our xG was only 1.28.

It's just arbitrary nonsense.

lol

Yesterday was a very good example of why it's not nonsense. First half, we had 4 shots on target and scored 4 goals, that doesn't happen very often. A combination of good finishing and poor keeping caused that. If you replay that first half over and over most of the time we would not score 4 goals.

 

[stupot07]

lol

Yesterday was a very good example of why it's not nonsense. We had 4 shots on target and scored 4 goals, That doesn't happen very often. A combination of good finishing and poor keeping caused that. If you replay that first half over and over most of the time we would not score 4 goals.

We had 9 shots on target.

 

[skybluecam]

We had 9 shots on target.

Was talking about the first half.

The point is, if you look at the shots we took critically you would not expect us to score 4 goals. Sakamoto's - decent chance but gets scored less often than you'd think, and arguably should have been saved. Eccles 1st - shot from 20 yards, rarely goes in. EMC - good chance but probably still only goes in half the time. Eccles 2nd - again, ~20 yards out, it gets missed more than scored.

Other than that we didn't create any particularly good chances.

People massively overestimate how 'good' chances actually are.

 

[PVA]

xG is still a load of nonense. We battered Plymouth yesterday 4-0, yet our xG was only 1.28.

It's just arbitrary nonsense.

It is, by definition, the very opposite of arbitrary.

 

[Ring Of Steel]

xG is still a load of nonense. We battered Plymouth yesterday 4-0, yet our xG was only 1.28.

It's just arbitrary nonsense.

totally clueless

 

[stupot07]

totally clueless

We know you are, you don't have to announce it on here.

 

[Tomh111]

I know we're high, as in good. Which I know suggests unlucky but I don't believe the errors we make which do lead to the goals we concede are unlucky tbh.

I agree a small amount, I don't think it's a collective issue.

Goalkeeping has been an issue, but our lack of closing down on the edge of the area is a chronic problem.

Sent from my SM-S911B using Tapatalk

 

[stupot07]

Was talking about the first half.

The point is, if you look at the shots we took critically you would not expect us to score 4 goals. Sakamoto's - decent chance but gets scored less often than you'd think, and arguably should have been saved. Eccles 1st - shot from 20 yards, rarely goes in. EMC - good chance but probably still only goes in half the time. Eccles 2nd - again, ~20 yards out, it gets missed more than scored.

Other than that we didn't create any particularly good chances.

People massively overestimate how 'good' chances actually are.

But that's football. We use xG to say we are better than we are, We use xGA to say we're not as bad as we are. But the reality is, players score goals, defenders and keepers make mistakes.

Lampard made a living our of scoring worldies, same as Le Tissier, Alan Shearer used to regular score screamers from outside the box. Theres only 1 stat that matters in football, and that's the score. Also unless you have some kind of Hawkeye technology monitoring the pace the trajectory, the angle of the ball, the exact position of everyone on the pitch, where the keeper is, the angle you're approaching it, players height, pace of the place traveling onto a ball, the wind speed and direction, it's complete nonsense. Even the type of ball, the EFL balls travel differently to the PL.

In fact xG would differ country to country dependent on the air pressure, humidity, etc.

Football is random, it's why we love it.

 

[Ring Of Steel]

But that's football. We use xG to say we are better than we are, We use xGA to say we're not as bad as we are. But the reality is, players score goals, defenders and keepers make mistakes.

Lampard made a living our of scoring worldies, same as Le Tissier, Alan Shearer used to regular score screamers from outside the box. Theres only 1 stat that matters in football, and that's the score. Also unless you have some kind of Hawkeye technology monitoring the pace the trajectory, the angle of the ball, the exact position of everyone on the pitch, where the keeper is, the angle you're approaching it, players height, pace of the place traveling onto a ball, the wind speed and direction, it's complete nonsense. Even the type of ball, the EFL balls travel differently to the PL.

maybe try to understand it more.

 

[skybluecam]

But that's football. We use xG to say we are better than we are, We use xGA to say we're not as bad as we are. But the reality is, players score goals, defenders and keepers make mistakes.

Lampard made a living our of scoring worldies, same as Le Tissier, Alan Shearer used to regular score screamers from outside the box. Theres only 1 stat that matters in football, and that's the score. Also unless you have some kind of Hawkeye technology monitoring the pace the trajectory, the angle of the ball, the exact position of everyone on the pitch, where the keeper is, the angle you're approaching it, players height, pace of the place traveling onto a ball, the wind speed and direction, it's complete nonsense. Even the type of ball, the EFL balls travel differently to the PL.

You're letting the perfect be the enemy of the good.

"There's only 1 stat that matters and that's the score" - for determining the result yes, for analysing performance, no. Anyone who knows football will realise it's massively affected by variance. You can easily be the better team and end up losing.

 

[PVA]

Lampard made a living our of scoring worldies, same as Le Tissier, Alan Shearer used to regular score screamers from outside the box.

Well yes, exactly.

That's why they were world class players - they were able to consistently score incredibly difficult chances/score out of nothing

ie they would massively outperform their xG stats which would show them as incredibly valuable players.

You've just described one use of xG without realising

 

[Gosb]

On the plus side, our 32 league goals so far (5th best in the Division) have been scored by 13 different players.

 

[rob9872]

Nonsense or not, if Collins had let in the 4 their keeper did yesterday, the fall out on here (including by me) would've been toxic. He was shocking and however good we were, another day there's no way we get those 4 again (akrgough should've had a penalty so 1.28 does seem low)

 

[sixfoursix]

Nonsense or not, if Collins had let in the 4 their keeper did yesterday, the fall out on here (including by me) would've been toxic. He was shocking and however good we were, another day there's no way we get those 4 again (akrgough should've had a penalty so 1.28 does seem low)

I don't think a non-awarded "penalty" counts towards xG

 

[wingy]

I thought Tats goal was a Tremendous leap Speedieesque but unchallenged, where was the defence?

 

[rob9872]

I thought Tats goal was a Tremendous leap Speedieesque but unchallenged, where was the defence?

I liked how it was something we've been working on withe wide players getting in on the opposite flank. Lampard mentioned it with Ephrons goal a couple of weeks back. Small changes but nice to see that progress.

 

[Tomh111]

Well yes, exactly.

That's why they were world class players - they were able to consistently score incredibly difficult chances/score out of nothing

ie they would massively outperform their xG stats which would show them as incredibly valuable players.

You've just described one use of xG without realising

Only one footballer has ever outscored xG every season of their career, that's Messi.

*That was a few years ago so may have changed*

But the premises of xG is that like all stats, eventually they regress tk the mean, no player routinely/consistently outscored the average.

There's no such thing as a better finisher, just people in hot streaks and those that aren't.

It's one of the important parts of using stats, to avoid confirmation bias. Because we all remember the one that goes in and not the 45 that miss, the acca that wins not the 75 that don't and the one Rochard Shaw worldy and not he 100 Konjic misses.

Sent from my SM-S911B using Tapatalk

 

[Tomh111]

I thought Tats goal was a Tremendous leap Speedieesque but unchallenged, where was the defence?

Tats seems to win far more headers than someone of his stature should, he reads the ball flight very well and seems to get up at the right moment.

Sent from my SM-S911B using Tapatalk

 

[SlowerThanPlatt]

Ever walked away from a game and said “we should have won today” or “we got away with that”?

xG is just a statistical model to back that up removing bias.

 

[Philosoraptor]

Ever walked away from a game and said “we should have won today” or “we got away with that”?

xG is just a statistical model to back that up removing bias.

A good analogy of the model is someone taking a few spray cans to a car that has been in a huge prang, and saying; 'that will sort it out'.

 

[Otis]

I know we're high, as in good. Which I know suggests unlucky but I don't believe the errors we make which do lead to the goals we concede are unlucky tbh.

I think where we are perhaps a little bit unlucky, is that pretty much every mistake at the back, leads to an opposition goal.

Usually you can make mistakes and get away with it on many occasion, because the opposing player makes a hash of things and you get a big let-off.

That has of course happened, but many a time we have made just one awful mistake and it's been a goal and/or two awful mistakes and it's been two goals.

Did start to think we were jinxed at some point, but I think it happens to many a team and you can go through spells where every error at the back results in a goal

 

[skybluecam]

More stats for you all

Our opponents converted that 23.0 xGA into 25.3 PSxG - suggesting weve faced slightly better than average finishing, nothing crazy though.

From that PSxG we've conceded 34, 33 discounting an own goal. That would suggest ~7.7 of the goals we've conceded are down to the keeper.

Breaking that down by keeper you get

Dovin -0.4 PSxG (-0.04 p/90) Wilson -1.1 PSxG (-0.36 p/90) Collins -6.1 PSxG (-0.77 p/90)

So if we'd just stuck with Dovin you could estimate we would have conceded 6 or 7 less goals.

 

[robbiekeane]

Well yes, exactly.

That's why they were world class players - they were able to consistently score incredibly difficult chances/score out of nothing

ie they would massively outperform their xG stats which would show them as incredibly valuable players.

You've just described one use of xG without realising

I might print this out and frame it above my bed

 

[clint van damme]

I might print this out and frame it above my bed

Good idea, you wouldn't want anything up there that was a passion killer.

 

[clint van damme]

View attachment 40436

Fair play. I'm a big admirer of a man who can simultaneously keep his Bluetooth permanently on and maintain a healthy charge in his phone.

 

[fernandopartridge]

xG is still a load of nonense. We battered Plymouth yesterday 4-0, yet our xG was only 1.28.

It's just arbitrary nonsense.

Fotmob saying it was 2.0 xG to 0.25 for Plymouth. That's showing complete dominance really, an xG eight times that of the opposition.

 

[stupot07]

Fotmob saying it was 2.0 xG to 0.25 for Plymouth. That's showing complete dominance really, an xG eight times that of the opposition.

Funny they have obviously changed it because footmob had it as this when I looked after the game. So according to the xG, 1-0 would have been a fair result based on quality of chances created.