xG

B

BoroPhil

Well-known member
For everyone who loves xG.

That was only the 2nd time we have 'beaten' the xG to win a league game this season (the other was Swansea away)

On the other hand we've lost when having higher xG 6 times (Coventry A, West Brom A, QPR H, Hull H, Rotherham A, Bristol C H)

We also had the higher xG in 4 of our 5 draws (Sheff Wed A the only exception)

And yesterday was our 3rd highest xG conceded of the season (after Blackburn A and Leeds A)

As well as our 2nd lowest xG of the season (only Ipswich H was worse)
 
BoroPhil said:
For everyone who loves xG.

That was only the 2nd time we have 'beaten' the xG to win a league game this season (the other was Swansea away)

On the other hand we've lost when having higher xG 6 times (Coventry A, West Brom A, QPR H, Hull H, Rotherham A, Bristol C H)

We also had the higher xG in 4 of our 5 draws (Sheff Wed A the only exception)

And yesterday was our 3rd highest xG conceded of the season (after Blackburn A and Leeds A)

As well as our 2nd lowest xG of the season (only Ipswich H was worse)
Click to expand...
That’s quite encouraging really .

So if results were decided on xG rather than goals .
If my thinking is correct ……
We would have 19 more points (4 x 3 from wins and 4 x 2 from the draws -less one for the Wednesday game . )
Of course draws will always happen - but we
could easily have had 10 more points and be in the playoff places as things stand ( even withstanding the battering we have taken with injuries )
 
At the other end of the scale Preston have defied the xG gods by winning NINE times whilst behind on xG (though not against us) and not lost a single game with higher xG

In fact they've only been ahead in 8 games compared to our 20 yet they have 14 wins to our 13
 
Talk me through this as I finally want to understand. Is it a set figure worked out before the match based on chances / goals in recent previous games. Or is it a moving metric through the current match only reacting to what is happening. If so , when does it move , by the 5 min , or only on goal/near goal events, and also if so, at kick off it would be 0 right ? as the ball hasnt gone near the goal.
 
Boro_Interlinked said:
Talk me through this as I finally want to understand. Is it a set figure worked out before the match based on chances / goals in recent previous games. Or is it a moving metric through the current match only reacting to what is happening. If so , when does it move , by the 5 min , or only on goal/near goal events, and also if so, at kick off it would be 0 right ? as the ball hasnt gone near the goal.
Click to expand...
Yeah your last point is correct. It starts at zero at kick off and every shot/chance is allocated an xG rating which is from 0-1 depending on the perceived difficulty
 
Boro_Interlinked said:
Talk me through this as I finally want to understand. Is it a set figure worked out before the match based on chances / goals in recent previous games. Or is it a moving metric through the current match only reacting to what is happening. If so , when does it move , by the 5 min , or only on goal/near goal events, and also if so, at kick off it would be 0 right ? as the ball hasnt gone near the goal.
Click to expand...
Yes to all the above.

Teams have an historic xg based on previous games.

Plenty of stat sites will show you a moving xg whilst in-play. The delay in-play is based on when the stats update. Some apps you can refresh when you like, fotmob for example.

I would always be very cautious of using xg on its own for any stats based analysis. It's the only stat of it's kind
 
Jedi boro said:
Who created this and when did it appear as it seems to be everywhere now.
Click to expand...
Opta created it Jedi. Their first version was 2018 as I recall but the start was first described in 2014 by, and don't quote me Sam Greenwood.

Each of the three companies who calculate xg use a slightly different mechanism to calculate xg and even within companies the methodology has changed over time.

Opta is generally considered the best. It uses ai to give a consistent measure. Consistent doesn't mean accurate in this context. It just removes a lot of the human judgement.
 
Ok, ta . So one last point. Is it cumulative then and presumably can't go down . Like if your team got it to 0.5 but then didn't even have control of the ball for 15min period would it just stay at what it was before the 15m period or would they shave a bit off because of no likelihood of chances..
 
Boro_Interlinked said:
Ok, ta . So one last point. Is it cumulative then and presumably can't go down . Like if your team got it to 0.5 but then didn't even have control of the ball for 15min period would it just stay at what it was before the 15m period or would they shave a bit off because of no likelihood of chances..
Click to expand...
Yeah exactly. It's the cumulative total of all your chances during the game.
 
Boro_Interlinked said:
Ok, ta . So one last point. Is it cumulative then and presumably can't go down . Like if your team got it to 0.5 but then didn't even have control of the ball for 15min period would it just stay at what it was before the 15m period or would they shave a bit off because of no likelihood of chances..
Click to expand...
Over 90 minutes it goes up. Over a season it can go up and down because it is a measure normalised over 90 minutes.
 
I do think its useful to have a measure that measures a team's overall performance as we all know results don't allows reflect performance.
 
Redwurzel said:
I do think its useful to have a measure that measures a team's overall performance as we all know results don't allows reflect performance.
Click to expand...
I loathe it as a statistic. Largely because people tend to take it at face value. You are right it's a useful measure as are all footie stats. Over a season they usefuk, over 90 minutes they are woeful.
 
You must log in or register to reply here.
Back
Top