Expected Goals. There and back again

[RobinHood]

I'm old fashioned - only interested in actual goals and assists. Expected wins nothing.

 

[I'm With Stupid]

I'm old fashioned - only interested in actual goals and assists. Expected wins nothing.

Yeah, basically a few years ago, these fancy data analytics companies started making themselves available to regular punters, and what you've got now is a bunch of journalists and Youtubers who think they're experts because they can support what they already think with a bunch of out-of-context stats. So you get shit like the highest expected goals from outside the box with his left foot on a weekday, and basically you can cherry pick the stats to make every player look amazing in the same way that Youtuber compilations just show you their best moments.

I think one of the biggest clues that you'd bought a dud is when you get shown a list of quite obscure stats from someone telling you how good a player is. Meanwhile, someone like Reijnders is voted midfielder of the season in Serie A, and surprise, surprise, he looks quality. I've got no idea what his stats are, because no-one's trying to polish a turd with him.

 

[Vienna_70]

Yeah, basically a few years ago, these fancy data analytics companies started making themselves available to regular punters, and what you've got now is a bunch of journalists and Youtubers who think they're experts because they can support what they already think with a bunch of out-of-context stats. So you get shit like the highest expected goals from outside the box with his left foot on a weekday, and basically you can cherry pick the stats to make every player look amazing in the same way that Youtuber compilations just show you their best moments.

I think one of the biggest clues that you'd bought a dud is when you get shown a list of quite obscure stats from someone telling you how good a player is. Meanwhile, someone like Reijnders is voted midfielder of the season in Serie A, and surprise, surprise, he looks quality. I've got no idea what his stats are, because no-one's trying to polish a turd with him.

I’ve said, ever since it became a thing, that xGoals is a load of bollocks.

 

[bluesmith]

I’ve said, ever since it became a thing, that xGoals is a load of bollocks.

Its not though is it. It literally means if you get players in better positions they are more likely to score goals. It was one of the main things City used to turn Sterling in to a goalscorer.

 

[RobinHood]

I’ve said, ever since it became a thing, that xGoals is a load of bollocks.

Its not though is it. It literally means if you get players in better positions they are more likely to score goals. It was one of the main things City used to turn Sterling in to a goalscorer.

I think it's bollocks during a game, as xG doesn’t reflect the actual score line (which is what wins games). It’s a diagnostic tool.

 

[bluesmith]

I think it's bollocks during a game, as xG doesn’t reflect the actual score line (which is what wins games). It’s a diagnostic tool.

Its not there to tell you who's winning the game, it's there to give an indication in who's playing better, just like, shots on target, corner, possession etc.

 

[RobinHood]

Its not there to tell you who's winning the game, it's there to give an indication in who's playing better, just like, shots on target, corner, possession etc.

1) Actual goals shows who's winning.

2) xGoals indicates the probability a shot becomes a goal, based on its estimated difficulty. Meh!

 

[bluesmith]

1) Actual goals shows who's winning.

2) xGoals indicates the probability a shot becomes a goal, based on its estimated difficulty. Meh!

If you dont think a stat that tells you how good a teams chances are is important that fine, City and every top team do, and use it to improve players.

 

[bobbyowenquiff]

Its not there to tell you who's winning the game, it's there to give an indication in who's playing better, just like, shots on target, corner, possession etc.

It’s really just a measure of build-up play which is a useful tool as part of a bigger picture.

 

[bluesmith]

It’s really just a measure of build-up play which is a useful tool as part of a bigger picture.

You'll have to expand on that, as far as I'm aware it's based solely on the position a shot is taken from.

 

[bobbyowenquiff]

You'll have to expand on that, as far as I'm aware it's based solely on the position a shot is taken from.

It shows how many openings are created whether or not the chance is finished.

 

[bluesmith]

It shows how many openings are created whether or not the chance is finished.

It doesnt take into account any build up play just the position the shot is taken from

 

[Donmcfc]

Its not there to tell you who's winning the game, it's there to give an indication in who's playing better, just like, shots on target, corner, possession etc.

Why don't we just use shots on target, corners, possession etc.

 

[bluesmith]

Why don't we just use shots on target, corners, possession etc.

Because having one shot on target doesn't tell us much it could of been from 40 yards or 4 yards. Having one shot on target and an xG of 0.8 tells me we have missed a sitter.

 

[Donmcfc]

Because having one shot on target doesn't tell us much it could of been from 40 yards or 4 yards. Having one shot on target and an xG of 0.8 tells me we have missed a sitter.

So do we get 0.8 of a goal? No we don't. A miss is a miss is a miss. How do you account for the games when City have 75% possession but we lose? It's just trendy meaningless nonsense.
Keep statistics out of football except for the final score and points on the board because nothing else matters

 

[bluesmith]

So do we get 0.8 of a goal? No we don't. A miss is a miss is a miss. How do you account for the games when City have 75% possession but we lose? It's just trendy meaningless nonsense.
Keep statistics out of football except for the final score and points on the board because nothing else matters

It isn't meaningless though, City credit it as one of the main reasons why Sterling went from a goal every six games to a better than a goal every two games.

They also used it with Rodri positioning by telling him to get an extra 5 yards forward before shooting so he was on the edge of the box - how did that work out for us?

I use them for betting, if a team is consistently getting higher xG than there opponent they are more likely to win, finding teams with high xG who have lost a few on the trot can give false odds particularly if they have low shots on goal stats. Teams playing each other where they both have high xG are good options for BTTS bets etc.


If stats aren't your thing thats not a problem but don't call them meaningless because you dont understand the benefit of them.

 

[Donmcfc]

It isn't meaningless though, City credit it as one of the main reasons why Sterling went from a goal every six games to a better than a goal every two games.

They also used it with Rodri positioning by telling him to get an extra 5 yards forward before shooting so he was on the edge of the box - how did that work out for us?

I use them for betting, if a team is consistently getting higher xG than there opponent they are more likely to win, finding teams with high xG who have lost a few on the trot can give false odds particularly if they have low shots on goal stats. Teams playing each other where they both have high xG are good options for BTTS bets etc.


If stats aren't your thing thats not a problem but don't call them meaningless because you dont understand the benefit of them.

You really think Pep Guardiola needs bollocks like this to inform his players of where to stand. I presume you are a professional gambler with a million in the bank?
In the last two seasons I have seen so many games where City have totally dominated play in all areas of the pitch but have ended up as the losing team. Explain that. If xG is as good as you say how come we have just had our worst season in ten years. It's made us worse not better. It's absolute bollocks.

 

[cheekybids]

Because having one shot on target doesn't tell us much it could of been from 40 yards or 4 yards. Having one shot on target and an xG of 0.8 tells me we have missed a sitter.

Does XG distinguish brilliance. We have players who score worldies & then the XG is low but they are world class players. Is it a higher XG if the shot from R/H side is from Foden or Walker?

 

[bluesmith]

You really think Pep Guardiola needs bollocks like this to inform his players of where to stand. I presume you are a professional gambler with a million in the bank?
In the last two seasons I have seen so many games where City have totally dominated play in all areas of the pitch but have ended up as the losing team. Explain that. If xG is as good as you say how come we have just had our worst season in ten years. It's made us worse not better. It's absolute bollocks.

Well i've literally been at the CFA with City coaching staff and head of analytics who put on a full presentation showing the slides they went through with Sterling and explained how they used xG to improve his goalscoring. But Don off the internet knows more than the City coaches and analysts.

@twosips was at the same presentation.

 

[Donmcfc]

I have been watching football for over 60 years and I know that xG is bollocks. Why can't you explain how City had such a bad season and why half if the coaching staff have left the club. And just exactly where is Sterling now? He should be consistently one of the Premier League's top scorers if his xG indoctrination was worth a carrot. And when did a head of analytics ever lift a major trophy?