The xG nonsense

[MancMackem2]

I can't wait until the boffins finally get this xG etc cracked and we can all stop going to matches and buying Sky. FotMob can just release everyone's results for the week at 3pm on a Saturday.

Our lass will be delighted.

 

[SuperKevsLeftBoot]

Apologies I got the angle wrong, however watch it back, the moment he kicks the shot Fofana is literally pushing him over. That drops the xG to 0.33 over what would be higher with no defenders around.View attachment 27605

He’s nearly on top of the 6 yard line. It’s harder to miss that score that! If you don’t think that after looking at that image then there’s no hope.

 

[tunstallhill]

The fact is xG isn’t subjective it’s a mathematical calculation. That’s what people don’t get

thats fine i guess you'd have to deep dive into the intricacy of it all to argue if a chance was worth 0.1 or 0.5 ... but as long as you used the same scale for every opportunity, the results would be the same in the sense that one team would have more xG than the other, by a set degree.

 

[SuperKevsLeftBoot]

A good analysis here, Mukiele’s stats are mind blowing.

You must be logged on to see media items

Next season is going to be shite then according to that. Can’t wait

 

[Charmless Man]

A lad who worked at city was on SSN yesterday discussing it, said Pep wouldn’t entertain the notion that xg was a thing.

Ok.

But it is a thing, and I think his sides tended to play in a way which maximised xG per shot. Cut backs from the byline to players in close range of the goal for example.

So he may well have espoused in it but he did maximise it for his team tactically.

 

[A Visiting Villan]

It has Aston Villa 13th and Sunderland 18th. 9 and 11 places out respectively.

It has Arsenal, Man United, Manchester City and Nottingham Forest in their exact positions. It has Liverpool, Bournemouth, Brentford, Fulham, Everton, West Ham, Wolves and Burnley as 1 place out from where they finished. The last two would have been exact positions too if Burnley hadn't equalised againat Wolves on Sunday.

But you've decided that it's useless based on a couple of outliers.

No. I've decided XG is useless based on watching my own team totally and utterly destroy it all season.

 

[MadMackem7]

He’s nearly on top of the 6 yard line. It’s harder to miss that score that! If you don’t think that after looking at that image then there’s no hope.

Yeah but do you think it’s easier to score a chance when your body is fully balanced or being bent over by a centre back?

No. I've decided XG is useless based on watching my own team totally and utterly destroy it all season.

You scored long shot after long shot mate, low xG shots.

 

[SuperKevsLeftBoot]

Apologies I got the angle wrong, however watch it back, the moment he kicks the shot Fofana is literally pushing him over. That drops the xG to 0.33 over what would be higher with no defenders around.View attachment 27605

Snap shot reijnders chance against city is 0.40 how? Far tighter angle .. this is why I think it’s bllx.

100,000 shots by who? Where is the data behind it?
Someone has come up with XG and it’s pumped all over the media without any context to how the data behind the xg figure is derived. Could be 100,000 Milton Nunez shots for all we know.

 

[MadMackem7]

Snap shot reijnders chance against city is 0.40 how? Far tighter angle .. this is why I think it’s bllx.

100,000 shots by who? Where is the data behind it?
Someone has come up with XG and it’s pumped all over the media without any context to how the data behind the xg figure is derived. Could be 100,000 Milton Nunez shots for all we know.

Erm, quite clearly this is easier then brobbeys, the nearest defender is 3 yards behind rather than on Brobbeys back.

Also, the 100,000 shots is data pulled from matches, then the location of shots is marked with the type, the assisting pass, location of defenders then a calculation is done based on how many are scored

 

[SuperKevsLeftBoot]

View attachment 27606

Erm, quite clearly this is easier then brobbeys, the nearest defender is 3 yards behind rather than on Brobbeys back.

Also, the 100,000 shots is data pulled from matches, then the location of shots is marked with the type, the assisting pass, location of defenders then a calculation is done based on how many are scored

He doesn’t have as much of the goal to hit and btw your screen shot of brobbeys was too far ahead.

 

[A Visiting Villan]

Yeah but do you think it’s easier to score a chance when your body is fully balanced or being bent over by a centre back?

You scored long shot after long shot mate, low xG shots.

Yes because we have quality players who can do that. It doesn't mean we should be 9 places higher and that's what people read from the XG table.

 

[Morny]

He’s nearly on top of the 6 yard line. It’s harder to miss that score that! If you don’t think that after looking at that image then there’s no hope.

Off the top of my head we've had four players in a similar position this season.

Isidor Vs Villa
Diarra Vs Villa
Isidor Vs Liverpool
Brobbey Vs Chelsea

How many did we score?

 

[MadMackem7]

He doesn’t have as much of the goal to hit and btw your screen shot of brobbeys was too far ahead.

That one any better of him getting bent over?

Also yes it’s a tighter angle, which is why it’s only 0.4 xG despite having no defenders, if he was more central it would be higher, you’re starting to get the hang of this!

Off the top of my head we've had four players in a similar position this season.

Isidor Vs Villa
Diarra Vs Villa
Isidor Vs Liverpool
Brobbey Vs Chelsea

How many did we score?

Both chances vs Villa are 0.65 which is consistent as they were very similar

Vs Liverpool is harder as the chance is different when he took it round the keeper, xG is measured when the shot is taken which I can understand peoples frustration

 

[SuperKevsLeftBoot]

Off the top of my head we've had four players in a similar position this season.

Isidor Vs Villa
Diarra Vs Villa
Isidor Vs Liverpool
Brobbey Vs Chelsea

How many did we score?

Similar is vague. Isidore he scored but was more difficult until Martinez slip. Further out and the keeper was further out.
Diarra was More difficult for the same reason
Isidors v Liverpool was too straight.
Brobbey is 8 yards out at a good angle and yes a defender is on him but there is so much space in the bottom right… which is where most strikers would have put it.
I’m not digging brobbey but his finishing is rash. Even the second goal his shot was going wide

 

[Charmless Man]

View attachment 27607

That one any better of him getting bent over?

Also yes it’s a tighter angle, which is why it’s only 0.4 xG despite having no defenders, if he was more central it would be higher, you’re starting to get the hang of this!

Both chances vs Villa are 0.65 which is consistent as they were very similar

Vs Liverpool is harder as the chance is different when he took it round the keeper, xG is measured when the shot is taken which I can understand peoples frustration

Don’t think people realise a penalty is only ~0.75-0.76 xG.

If you have the keeper closing the angle a defender on your back and you’re off balance (which I think most models take account of) it’s going to be substantially lower.

 

[SuperKevsLeftBoot]

View attachment 27607

That one any better of him getting bent over?

Also yes it’s a tighter angle, which is why it’s only 0.4 xG despite having no defenders, if he was more central it would be higher, you’re starting to get the hang of this!

Both chances vs Villa are 0.65 which is consistent as they were very similar

Vs Liverpool is harder as the chance is different when he took it round the keeper, xG is measured when the shot is taken which I can understand peoples frustration

0.65 hahaha we’ll leave it there. My point proven cheers

 

[MadMackem7]

0.65 hahaha we’ll leave it there. My point proven cheers

They were free shots at goal inside the box. Penalties are around 0.79. I think 0.65 is very fair.

Don’t think people realise a penalty is only ~0.75-0.76 xG.

If you have the keeper closing the angle a defender on your back and you’re off balance (which I think most models take account of) it’s going to be substantially lower.

I don’t understand, the fella has now said Isidor and Diarra vs Villa, and that reijnders chance I posted were harder than brobbeys

They all had free shots on goals, Brobbey is getting pushed from behind

 

[Morny]

The problem for xG is that it is so often used (or criticised) by incredibly partisan fans. For example, the fans on here who HATE XG IT'S JUST STUPID because we perform poorly on it. The flip side is (for example) Leeds fans on twitter, who go on about it all the time because they performed better on it.

So, like everything in football nowadays, so much of the discourse about it is coloured by people who are so swept up in their own biases that it's pointless arguing.

Facts are:

• It isn't perfect, and doesn't tell a full story at all.
• It is relatively useful, to get an at a glance story of how a game's was in terms of chances. If I'm out and want to check a score in a pinch, I know if i'll check it on a website with xG I can get a *very broad* idea of how the play has been.
• It is used by professionals to some degree; betting companies have been using it (and more sophisticated variations) for years, and still do (see our odds for next season, they expect regression). Football clubs use it (and more sophisticated variations) to assist with player recruitment/identification, hand-in-hand with old-fashioned scouting.
• Things like xG tables are just a bit of fluff, same as "tables by goals scored after 90 mins" or "second half tables" or whatever. People take them far too serious (on both sides of the debate) and should just chill out. It's like being absolutely livid because the "league table if you remove Var changes" table has you in a different position. It means nowt.
  

View attachment 27607

That one any better of him getting bent over?

Also yes it’s a tighter angle, which is why it’s only 0.4 xG despite having no defenders, if he was more central it would be higher, you’re starting to get the hang of this!

Both chances vs Villa are 0.65 which is consistent as they were very similar

Vs Liverpool is harder as the chance is different when he took it round the keeper, xG is measured when the shot is taken which I can understand peoples frustration

I know they're different, it was just a point highlighting that all chances deemed "easy" aren't actually easy. All four of them, people would see them and think "that's an easy chance, players should always score".

 

[MadMackem7]

The problem for xG is that it is so often used (or criticised) by incredibly partisan fans. For example, the fans on here who HATE XG IT'S JUST STUPID because we perform poorly on it. The flip side is (for example) Leeds fans on twitter, who go on about it all the time because they performed better on it.

So, like everything in football nowadays, so much of the discourse about it is coloured by people who are so swept up in their own biases that it's pointless arguing.

Facts are:

• It isn't perfect, and doesn't tell a full story at all.
• It is relatively useful, to get an at a glance story of how a game's was in terms of chances. If I'm out and want to check a score in a pinch, I know if i'll check it on a website with xG I can get a *very broad* idea of how the play has been.
• It is used by professionals to some degree; betting companies have been using it (and more sophisticated variations) for years, and still do (see our odds for next season, they expect regression). Football clubs use it (and more sophisticated variations) to assist with player recruitment/identification, hand-in-hand with old-fashioned scouting.
• Things like xG tables are just a bit of fluff, same as "tables by goals scored after 90 mins" or "second half tables" or whatever. People take them far too serious (on both sides of the debate) and should just chill out. It's like being absolutely livid because the "league table if you remove Var changes" table has you in a different position. It means nowt.

Spot on mate, xG tables are a bit of a stretch from just xG, they get stuff right but it isn’t the be all and end all

The problem for xG is that it is so often used (or criticised) by incredibly partisan fans. For example, the fans on here who HATE XG IT'S JUST STUPID because we perform poorly on it. The flip side is (for example) Leeds fans on twitter, who go on about it all the time because they performed better on it.

So, like everything in football nowadays, so much of the discourse about it is coloured by people who are so swept up in their own biases that it's pointless arguing.

Facts are:

• It isn't perfect, and doesn't tell a full story at all.
• It is relatively useful, to get an at a glance story of how a game's was in terms of chances. If I'm out and want to check a score in a pinch, I know if i'll check it on a website with xG I can get a *very broad* idea of how the play has been.
• It is used by professionals to some degree; betting companies have been using it (and more sophisticated variations) for years, and still do (see our odds for next season, they expect regression). Football clubs use it (and more sophisticated variations) to assist with player recruitment/identification, hand-in-hand with old-fashioned scouting.
• Things like xG tables are just a bit of fluff, same as "tables by goals scored after 90 mins" or "second half tables" or whatever. People take them far too serious (on both sides of the debate) and should just chill out. It's like being absolutely livid because the "league table if you remove Var changes" table has you in a different position. It means nowt.

I know they're different, it was just a point highlighting that all chances deemed "easy" aren't actually easy. All four of them, people would see them and think "that's an easy chance, players should always score".

Oh yeah my point is just, the person debating with me clearly seems to suggest distance from goal = easiness of the chance. Which just isn’t the full case

 

[SuperKevsLeftBoot]

They were free shots at goal inside the box. Penalties are around 0.79. I think 0.65 is very fair.

I don’t understand, the fella has now said Isidor and Diarra vs Villa, and that reijnders chance I posted were harder than brobbeys

They all had free shots on goals, Brobbey is getting pushed from behind

Diarra’s chance is harder than brobbey’s. The other issue is that you’re not seeing the goal from the attackers view. Diarra will see little of the goal if he looked up hence going for the chip. Brobbey (not sure he did look up) will see a boat load of space on the right. He should easily have finished it like wilson. He’s also so much closer whether being pushed or not!