Does xG tell us anything about the season?

[Silvio Dante]

Green line - non penalty xG for over rolling 6 games. Red line - non penalty xG against over rolling 6 games:

I’d say it’s got a pretty good correlation (NB - I take xGs value as collective as opposed to individual chance level as being a decent arbiter)

 

And here’s the source (and everyone else)

 

[Numero Uno]

If XG was the be all and end all Ipswich would be ****** and Watford would be right up there.

[BigTone]

I must admit to finding this stuff to be a complete load of bollox. You can have XG, MG, TV, D&G, XTC, MTV, E TYPE etc etc but at the end of the day you need to score more goals than the opposition and win games.

Edited April 28 by BigTone

[Numero Uno]

5 minutes ago, Numero Uno said:

If XG was the be all and end all Ipswich would be ****** and Watford would be right up there.

Apologies I was reading above the name not below

[TV Tom]

No

[ZiderEyed]

Think I've had my fill of stats in football. Makes me sick looking at lines and charts all the time trying to unearth some speck of hope for us. 

[Kid in the Riot]

No. Put down the graphs and spreadsheets and enjoy this lovely April evening. 

[Davefevs]

@Silvio Dante yes!  When you take it as a 6-game rolling average / season, it can highlight some trends.  Is it the be-all and end-all, nope.  But nobody is saying it is all-defining are they, either?

[IAmNick]

Thread: does anyone find this chart interesting?

Answer: stats in football make me physically sick and trying to find any meaning in anything beyond a result is pointless why bother life is hopeless and meaningless anyway so let's all just stop even trying

Overreaction much!

[ExiledAjax]

Yes xG can tell us things. It's not always new things, but it is things.

Hasn't failed me for about 6 years now.

[Tim Monaghan]

xG is the most nonsense stat in football. I also see the nonsense stat thread, that was updated after every match to batter Manning has disappeared…… funny that. 

Edited April 28 by Tim Monaghan

[Davefevs]

11 minutes ago, Tim Monaghan said:

xG is the most nonsense stat in football. I also see the nonsense stat thread, that was updated after every match to batter Manning has disappeared…… funny that. 

Come on Timmy, Mr P updated a post last night with a new table, a few people commented on it.

It’s such nonsense that it’s even been quoted by our beloved head-coach to defend performances.

 

 

[Tim Monaghan]

21 minutes ago, Davefevs said:

Come on Timmy, Mr P updated a post last night with a new table, a few people commented on it.

It’s such nonsense that it’s even been quoted by our beloved head-coach to defend performances.

 

 

Just because Manning has quoted it, doesn’t mean it’s not nonsense.

 

Edited April 28 by Tim Monaghan

[IAmNick]

1 hour ago, Tim Monaghan said:

xG is the most nonsense stat in football. I also see the nonsense stat thread, that was updated after every match to batter Manning has disappeared…… funny that. 

Why do you think it's nonsense?

[chinapig]

50 minutes ago, Tim Monaghan said:

Just because Manning has quoted it, doesn’t mean it’s not nonsense.

 

What about all the other coaches from Pep down who use analytics as one of the tools available to them (not just xG that's just the only one pundits can cope with).

Shall I tell them they've all got it wrong or will you?

[Rocking Red Cyril]

2 hours ago, BigTone said:

I must admit to finding this stuff to be a complete load of bollox. You can have XG, MG, TV, D&G, XTC, MTV, E TYPE etc etc but at the end of the day you need to score more goals than the opposition and win games.

Tone you spot on

[Arsene's Wanger]

Post an xg league table compared to the actual league table. Then you'll see if it tells us anything about the season.

[IAmNick]

6 minutes ago, Arsene's Wanger said:

Post an xg league table compared to the actual league table. Then you'll see if it tells us anything about the season.

This is xGD - so xG - xGA (expected goals minus expected goals against). If it was pure xG then it's just who you'd expect to score the most, which isn't necessarily who you'd expect to be top.

It's pretty accurate. Not perfect and obviously some outliers but definitely a correlation:

(there are multiple xG models obviously and this is just one)

[Mr Popodopolous]

2 hours ago, Tim Monaghan said:

xG is the most nonsense stat in football. I also see the nonsense stat thread, that was updated after every match to batter Manning has disappeared…… funny that. 

Pfft. Manning has got back to parity.

That is good and the recent trend is very encouraging and thst is broadly good.

I was expecting Manning to at last maintain the level and he has via a circuitous route and eventually. Well done. Brought some time, as well as some credit.

Credit where due, correct. If he has learned lessons all the better.

Edited April 28 by Mr Popodopolous

[Tim Monaghan]

37 minutes ago, chinapig said:

What about all the other coaches from Pep down who use analytics as one of the tools available to them (not just xG that's just the only one pundits can cope with).

Shall I tell them they've all got it wrong or will you?

Can you show me where I said all analytics? 

[Tim Monaghan]

3 minutes ago, IAmNick said:

This is xGD - so xG - xGA (expected goals minus expected goals against). If it was pure xG then it's just who you'd expect to score the most, which isn't necessarily who you'd expect to be top.

It's pretty accurate. Not perfect and obviously some outliers but definitely a correlation:

(there are multiple xG models obviously and this is just one)

Said xG. So show the xG table. That’s what we are talking about. So you’ve just proven my point that it’s nonsense. 

[MyBrotherErnie]

A huge problem in football media is that too many people use statistics without referring back to why they matter.

xG, in itself, obviously doesn't matter (at all). But it does measure something that does matter -- the quality of chances a team is creating. Too often journalists (and sometimes managers) get confused on this, and start to speak as if it is the statistic itself that matters (see this). The statistic is not useful unless it is good for measuring something meaningful!

In this case, of course it doesn't matter that our 6-game rolling average xG has improved, but this is good evidence to support the intuition that our quality of play has improved in the last month (beyond results), which does matter. 

[Tim Monaghan]

1 minute ago, MyBrotherErnie said:

A huge problem in football media is that too many people use statistics without referring back to why they matter.

xG, in itself, obviously doesn't matter (at all). But it does measure something that does matter -- the quality of chances a team is creating. Too often journalists (and sometimes managers) get confused on this, and start to speak as if it is the statistic itself that matters (see this). The statistic is not useful unless it is good for measuring something meaningful!

In this case, of course it doesn't matter that our 6-game rolling average xG has improved, but this is good evidence to support the intuition that our quality of play has improved in the last month (beyond results), which does matter. 

Exactly this. 

[IAmNick]

3 minutes ago, Tim Monaghan said:

Said xG. So show the xG table. That’s what we are talking about. So you’ve just proven my point that it’s nonsense. 

How does that prove your point? Here's the xG one, it's really similar:

 

Edited April 28 by IAmNick

[Davefevs]

9 minutes ago, MyBrotherErnie said:

A huge problem in football media is that too many people use statistics without referring back to why they matter.

xG, in itself, obviously doesn't matter (at all). But it does measure something that does matter -- the quality of chances a team is creating. Too often journalists (and sometimes managers) get confused on this, and start to speak as if it is the statistic itself that matters (see this). The statistic is not useful unless it is good for measuring something meaningful!

In this case, of course it doesn't matter that our 6-game rolling average xG has improved, but this is good evidence to support the intuition that our quality of play has improved in the last month (beyond results), which does matter. 

Nicely put, it’s not nonsense is very useful to try to measure a complex game, especially a low-scoring sport. XG along with other stuff…box entries (however much people scoff), etc, etc. is widely used, and when used beyond just in isolation.

Edited April 28 by Davefevs

[Tim Monaghan]

4 minutes ago, IAmNick said:

How does that prove your point? Here's the xG one, it's really similar:

 

To the current league table? It’s not though is it. 

[IAmNick]

3 minutes ago, MyBrotherErnie said:

A huge problem in football media is that too many people use statistics without referring back to why they matter.

xG, in itself, obviously doesn't matter (at all). But it does measure something that does matter -- the quality of chances a team is creating. Too often journalists (and sometimes managers) get confused on this, and start to speak as if it is the statistic itself that matters (see this). The statistic is not useful unless it is good for measuring something meaningful!

In this case, of course it doesn't matter that our 6-game rolling average xG has improved, but this is good evidence to support the intuition that our quality of play has improved in the last month (beyond results), which does matter. 

I disagree that Goodhart's Law is applicable, because xG can't be gamed in that way you're implying can it? It's an objective measure, so it's not like possession or shots on target - where you could play it around the back or shoot from 30 yards out to game the stats.

I think it does matter if something like an xG rolling average improves, because even if results don't match then it's an indication of where we're headed - and you'd expect in the majority of cases results to eventually converge with it. So it can be a leading indicator of improvement and can show an increase in performance that isn't yet paying dividends, but should be continued with, right?

[IAmNick]

4 minutes ago, Tim Monaghan said:

To the current league table? It’s not though is it. 

Isn't it?

The top 4 are the same top 4 in both tables.

8 of the top 10 are the same in both tables.

That's way, way above what you'd expect from random chance, so there's clearly a correlation. Is it perfect? Absolutely not - and nobody has ever said it is.

There are loads of papers on it's accuracy if you'd care to read them - which no doubt you won't, as you've already decided what you think!

 

Edited April 28 by IAmNick

[Tim Monaghan]

1 hour ago, IAmNick said:

Why do you think it's nonsense?

This is why I think it’s nonsense. Sometimes it’s Way Off. Way off. You’d think the team with the higher xG would win, right? Not always the case. It just measures the quality of chances, not who’s gonna win. 
Unlike what @Davefevs will have you believe, football is played by humans. Football isn’t played by robots; Players get nervous, they make mistakes, they do random stuff. xG doesn’t really get that.

Also xG is all about past data. But football’s always changing, so what happened before might not mean squat for what’s coming next. Injuries, suspensions, new players, change of tactics, the list is endless. 

So, yeah, xG is basic of the basic nonsense. 

[Tim Monaghan]

1 hour ago, chinapig said:

What about all the other coaches from Pep down who use analytics as one of the tools available to them (not just xG that's just the only one pundits can cope with).

Shall I tell them they've all got it wrong or will you?

All stats, yes. I’m not saying all stats are nonsense. But xG is IMO. 
 
P.s - “Statistics are just a pattern of information that we have. There are players who make the team play good without the statistics. You don’t need statistics. If you perform to your maximum, if you perform to your best and you helped your team-mates and make the process better, it is enough.” - Pep

 

 

Edited April 28 by Tim Monaghan