• The first stage of the forum upgrades has now been completed but they remain in a degraded state and are still being worked on. Normal posting/reading should now be possible.
    Please read this thread for more details.
    New user registrations are currently disabled.

This xG nonsense


I always think the penalty example illustrates the worth of xG quite well: penalties are scored approx 75% of the time, so they have an xG of 0.75. Now, you can’t score 3/4 of a penalty but if you take 4 penalties you should expect to score 3 of them based on general probability based on past events.

So with xG they’re trying to measure how good any chance is in the same way, based on a huge sample size of past events, accounting for position of shooter, type of assist, weak or strong foot etc (this is the difficult bit - trying to measure all the variables in chances that are more fluid than a penalty). The measurement also does not account for player ability in those moments, which is another weakness.

However to disregard it entirely seems a bit obtuse of people, and the fact that clubs use it must show that it tells us something.
 
jcrossan said:
If the Premier league table was set according to this xG silliness, that pundits are always mentioning and we are supposed to take seriously, we would be 4th bottom, just ahead of Aston Villa who would be 16th.

Thankfully we play on grass and not inside a computer program.
Click to expand...
You either score a goal or don't, simple as !
 
I think it's a useful metric. Nowhere near as useful as actual goals - obviously - but more useful than some of the stats we typically obsess over.

I grew up convinced that possession told you who then best team is, but then Leicester won the league with that season's 3rd worst possession. And always thought corners were another good indicator, but then I read that something like 2% of corners actually lead directly to a goal.
 
connolly_31 said:
In xG if the ball is rolled along the box and the striker just has a tap in, keeper is nowhere to be seen, open goal type affair but doesn’t actually touch the ball, does that go down as 0.00 xG? But if he blasted it owa or wide it would be something like 0.85 xG?
Click to expand...

That's the important question in all this, really. "What is a chance?"

Same goes for things like a player shaping to shoot on the edge of the box, changing his mind, taking a couple of touches then running into trouble and blasting something into a defender 5 seconds later.

What was the chance? Which has what XG and why? Which is used and why? Both are chances, both 'have' XG that could be attributed. Which is 'the' chance? It all says very different things about both the player and their attributes, depending on the answers to all the above.

Ultimately, like all stats, it's largely meaningless in isolation. The only people who really need to care about any of it are recruitment teams, and even then, theres a lot of guess work and massaging of stats going on. Most will default to watching with their eyes.

If, over 3 seasons, a player has an XG of 0.85 per 90, is he good or bad?
Same question for a team?

The answer is never in the stats, it's always in reality.

If the 0.85 player averages 1 goal in 3, he's likely half decent.
If the 0.85 team win 3 games all season, they are likely shite.
 
Last edited:
MiddleAgedMan said:
Do you not think a shot from edge of the area does only go in roughly 1 in 10?
Click to expand...

He shoots pretty much a foot from the penalty spot having taken a touch about three feet into the area.

Thus it makes measuring the chance he's taken a bit hard as he's received the ball on the edge of the box when running at pace. Knowing Brian Brobbey and his strengths we pretty much can suggest the chance of him having a good opportunity in that situation is higher than one in ten as that kind of opportunity is manna from heaven for him.

If its Enzo Le Fee he's probably taking the shot as soon as he receives the ball on the edge of the area, which makes it much harder or going to take a touch, take a step sideways and beat the defender in another way, thus meaning it's lesser chance. Had it been Enzo Le Fee then maybe the ball wouldn't have also been played in the same manner anyway.

So for me not taking into account the relative strength and weaknesses of the individual player in each situation means its lacking context and a bit shit as a result*



*though obviously that would make it even more complex and more open to question.
 
Binxy said:
Mansfield XG 0.86 - goals scored 4
Shef Utd XG 1.8 - goals scored 3

🤷‍♂️
Click to expand...
2 of Mansfields goals were long range. That’s 0.03 and 0.04 xG for both of them.

The other 2 were 1 in 3 chances.

As for Sheffield United, they’re now on 2.29 xG but one of their goals is an OG which xG doesn’t account for. So they’ve scored 2 goals from around 2 xG
 
You must log in or register to reply here.
Back
Top