Spurs to fade; Wolves are Ronseal

[Norman Deeley]

Interesting article in the Atheltic this morning. Looking at xG and the season so far. Main conclusions - Spurs will fade, Sheff United have been unlucky, Arsenal are not in a false position. And that the team in the Prem whose GF and GA most closely matches their xG is Wolves. Must say I'm quite impressed by the Athletic, although I'm currently on the quid a week offer, which seems about right.

 

[Tredman]

Interesting article in the Atheltic this morning. Looking at xG and the season so far. Main conclusions - Spurs will fade, Sheff United have been unlucky, Arsenal are not in a false position. And that the team in the Prem whose GF and GA most closely matches their xG is Wolves. Must say I'm quite impressed by the Athletic, although I'm currently on the quid a week offer, which seems about right.

Quid a week or quid a month..?

 

[Trips]

Quid a month the offer I'm on for 12 months

 

[AndyWolves]

I like xG as a stat, the only problem is deciding whether teams overperform the stat because they are better than everyone else or because they are lucky.

 

[MARKakaJIM]

It's a half decent indicator at a glance but you certainly need to delve a bit deeper into the individual chances and players involved to really make any use of it.

Think someone like Coutinho in his Liverpool days used to score a pretty high proportion of his goals from outside the box where xG would've been relatively low but it was a skill he'd honed and so I'd assume he'd have consistently outperformed from those situations. That overperformance would've carried over into the teams overall score but in reality they just had a player capable of executing those chances very well, broadbrush looks like overperformance but a detailed view shows that you couldn't afford to give up those chances around the edge of the box to that player.

I guess Sheff Utd are the opposite at the moment, they're still creating decent chances in most games but they're underperforming their xG, at a glance you put it down as bad luck and things maybe even themselves out but in reality their forward line is piss so there's a good chance they continue to spurn those chances.

Not sure what the sample size is for calculating the xG? Is it specific to certain leagues and years? Whatever it's averaged across the better players in the sample should invariably overperform it as they're above average and the weaker players should underperform it as they're the ones dragging the average down.

 

[YoungWolf]

That isn't really how it works.

Spurs have made an excellent habit of scoring from their first shot on target in so many games this season. This has allowed them to just defend (see - City and Arsenal games) which is why their xG is so low. So what will be interesting is when they don't score their first shot, because when teams have sat off against them they have tended to struggle. It can't last forever anyway. Son doesn't score that effort from that position many times out of a 100!

We haven't been great in in general so far but are slightly improving, and in the two years in the PL so far, we haven't tended to be a team that wildly over or underperforms, we are very reliable in that respect.

 

[YoungWolf]

Not sure what the sample size is for calculating the xG? Is it specific to certain leagues and years? Whatever it's averaged across the better players in the sample should invariably overperform it as they're above average and the weaker players should underperform it as they're the ones dragging the average down.

Statsbomb have the best and biggest sample size (and also cover the most variables) but that it probably the bit at the moment that is difficult, different companies have different algorithms too.

https://torvaney.github.io/projects/xG.html

This is a free one but it is very basic