Whats this xG stat all about?

Coopertrueblue

Coopertrueblue

Well-Known Member
Official Ticketer
Everton vs Chelsea

Everton 9 shots, Chelsea 17
Everton 4 on target, Chelsea 5
Everton 1.82 xG, Chelsea 0.84 xG

Not got a clue with that.
 
Expected goals.

Just a stat to measure the quality of chance that you have. The higher it is, the better type of chance you had.


I think it is useful to an extent. Wouldn't write it off as many like to do just because it is seen as nerd speak.
 
A penalty has an xG of about 0.8 because 4 out of every 5 are generally scored.

A tap in from a yard out will have an xG of like 0.99 as 99 times out 100 it will be a goal.

A 35 yard shot will have an xG of like 0.02 as 2 in 100 will be a goal.

It looks at the type of chance, the defenders between the player and the goal, whether it's a header or a shot, from a cross or a through ball, set piece etc...loads of factors.
 
Duncan Alexander, who I think was behind the stat, doesn't like the term "expected goals" - he prefers calling it "chance quality."
 
Coopertrueblue said:
Everton vs Chelsea

Everton 9 shots, Chelsea 17
Everton 4 on target, Chelsea 5
Everton 1.82 xG, Chelsea 0.84 xG

Not got a clue with that.
Click to expand...
Basically says Chelsea have lots of attempts on goal - but are poor at converting (like Rangers!).

Everton, with fewer chances, are more likely to score (must be the first time this season that has happened!).
 
It's absolutely BS. I followed a Twitter account for a while that posted xG alongside the actual score. Only once was it bang on, out of hundreds of games
 
20rangersultras07 said:
xG is estimated goals, tbh it’s a nothing stat.

xG: Borussia Dortmund 4.28 - 1.86 Bochum -
Possession: 63/37
FT score: 3-4
Click to expand...
It’s not a nothing stat. It gives an indication of the quality of chances. If you constantly underperforming then it’s a good metric or we’re improvement is needed.
 
Curious to see how the xTable that we were all taking the piss out of (me included) back at the start of the season when we were on top and they were dropping points left, right and centre, compares to the final outcome.
 
Last edited:
Imagine Team A took 10 shots from distance or awkward angles , based on solely on shot on goal you would expect them to score more goals than Team B if Team B had only 3 shots.

But that doesn't take in to account Team B's 3 shots might've been near open goals, a cut back blazed over the bar from 3y.

xG is a calculation made up of the quality of the chances and how often a goal would be expected to come from that chance.

So 10 easy saves for the keeper generates a lower xG than 3 sitters.

That's a very simplistic representation but it's my interpretation.
 
Hate it or love it this is the bread and butter of all data analysts employed by football teams (including ours - we recently advertised a role for another data analyst to join the backroom staff).

There are many platforms that analyse matches (FutMob, StatsBomb etc) and each have slightly different models and will generate different xG's for each team in a match.

This kind of analysis is much superior to "the eye test" particularly when trying to figure out if a player is worth buying (albeit the eye test is also used in conjunction with stats).

Good article summarising the increasing use of data in football: https://www.bbc.co.uk/news/business-56164159

Folk can call it pish all they like but it's only going to get more important as tech advances
 
The boys that run Rangers Review use expected goals to run a projected league table.

They analysed the difference in expected goals and actual goals scored.

From about September/October last year they had Celtic winning the league, even after we'd beaten them and gone top. Their points tally is going to be roughly the same as their projected points tally, as well.

Never really understood the ambivalence towards this - it's just another tool used in football to assess performance, no different to analysing possession or shots on goal.

The "virgin chat" is also funny - someone ask Matthew Benham at Brentford if he thinks it's a pile of shite :D
 
Souness86 said:
What are you talking about bang on? Do you mean the xG matched the actual scoreline, which it isn't meant to anyway?
Click to expand...
Yes
I really don't understand it's value as a stat. Hope does a shot get analysed? How do you grade a chance? Is a shot from 18 yards the same as if benzema takes it compared to harry McGuire?

It's a nonsense IMO
 
I think the point is to compare expected goals and actual goals scored. If xG is much higher than actual goals scored then your striker’s probably quite poor. If the opposite is true then the team is probably quite efficient with their attacks, like a good counter attacking team.
 
20rangersultras07 said:
xG is estimated goals, tbh it’s a nothing stat.

xG: Borussia Dortmund 4.28 - 1.86 Bochum -
Possession: 63/37
FT score: 3-4
Click to expand...

It’s not a nothing stat, it’s hugely important in how we sign players and utilise them, along with almost every modern football club. Advanced analytics charts were routinely mocked on here before Xmas as it had the mentally challengeds pipping us to the League as they were top of xG and xP tables. Ended up being correct.
 
It’s a gay football stat that makes Football Manager players feel like they really know about football.
 
It’s a pile of horseshit used to try and justify sides not getting a certain result.

It matters not one jot in the grand scheme of things but the younger generation will try to uphold it as in some way being meaningful.
 
Pacman said:
Yes
I really don't understand it's value as a stat. Hope does a shot get analysed? How do you grade a chance? Is a shot from 18 yards the same as if benzema takes it compared to harry McGuire?

It's a nonsense IMO
Click to expand...

The statistic measures the quality of chance being created, not the quality of the person taking the shot.

Here's an interesting thread about how virgins Brentford have used expected goals to reach the EPL:

https://twitter.com/x/status/1398232614201446401

I'm also surprised to see people write it off in a season where there's a direct link between our own expected goals and results.
 
I slaughtered this earlier in the season as it had Ross County in top 6 when they were bottom of the league....

It looks like it's a good indicator if teams are unlucky or on the right track and fortunes will turn and similarly if you are constantly winning by the odd goal and XG is really low then your likely to start dropping points.
 
ClockworkOrange said:
The statistic measures the quality of chance being created, not the quality of the person taking the shot.

Here's an interesting thread about how virgins Brentford have used expected goals to reach the EPL:

https://twitter.com/x/status/1398232614201446401

I'm also surprised to see people write it off in a season where there's a direct link between our own expected goals and results.
Click to expand...
That's exactly why I made the benzema/McGuire comment. A similar chance for both does not deserve the same expected outcome. Similarly a chance for the same player doesn't take into account whether it's on his better foot. Therefore it's a nonsense to say he has the same xG
 
Pacman said:
That's exactly why I made the benzema/McGuire comment. A similar chance for both does not deserve the same expected outcome. Similarly a chance for the same player doesn't take into account whether it's on his better foot. Therefore it's a nonsense to say he has the same xG
Click to expand...

I know what you're saying but it's less about the person taking the shot and more about the quality of chances that a team are creating.

If a team is creating a high xG score but the chances are falling to the wrong players, then that's probably indicative of something - players in the wrong positions, poor choices, etc. Similarly, if a team is overperforming on xG it maybe points towards players who are exceptionally good or a team who can't maintain their goal scoring. In short, it's used to analyse other things.

Like any other statistic, you need other numbers to make sense of it. No one looks at total number of shots or possession and thinks of it as being a definitive statistic. Expected goals is the same.
 
Fergie90 said:
I slaughtered this earlier in the season as it had Ross County in top 6 when they were bottom of the league....

It looks like it's a good indicator if teams are unlucky or on the right track and fortunes will turn and similarly if you are constantly winning by the odd goal and XG is really low then your likely to start dropping points.
Click to expand...

Same. I reckon one of the early season tables I took the piss out of will have turned out pretty accurate. Will post if I find one and hold my hands up.
 
Helps give you more of an idea of how the game actually went. You can have 20 shots from outside the box, but they aren’t expecting to go in.

Higher quality chance = higher xG.

The folk who think football is all about 442, big man and little man up top will hate it. But its useful to gauge a game.
 
You must log in or register to reply here.

Share this page

Latest posts

Back
Top