The Stats Thread

  • Please bear with us on the new site integration and fixing any known bugs over the coming days. If you can not log in please try resetting your password and check your spam box. If you have tried these steps and are still struggling email [email protected] with your username/registered email address
  • Log in now to remove adverts - no adverts at all to registered members!
Here's a similar comparison as a scatter graphic:

You must log in or register to see images
You must log in or register to see images
 
Last edited:
"Norwich’s rolling average for xG and xGa this season compared to 18/19. Early season attack is creating more, but giving away much more at the back. Recent trend is improving though." (Norwich City Stats)

There was quite a marked contrast in 2018--19 between the two halves of the season. In the first half, the first choice CB pairing was Zimmermann & Klose, with Leitner & Tettey, in front of them. Following the injuries to Klose and Leitner, the first choice CB pairing became Zimmermann & Godfrey, with McLean & Tettey in front of them. This change made us more potent in attack, but at the cost of being less solid in defence. No surprise that so far this season we are "creating more, but giving away much more". We have the same "main men" in attack as in the second half of 2018--19, while bedding in several newcomers in defence.
 
ncgandy said:
I love a stat, me. Better than having to sit through a boring 90 minutes. Loving this thread.
Click to expand...
There's no such thing as "a boring 90 minutes" of football. "Boringness" is in the eye (or rather, mind) of the beholder (or reader for that matter). Keep safe and have a bubbly Gandy Christmas <bubbly> <cracker> <bubbly>
 
  • Like
Reactions: ncgandy
Why we need Krul back, reason [HASHTAG]#366[/HASHTAG]:

You must log in or register to see media
 
Golden Eadie 2 said:
I have an unknown stat for you all. 87.68% of all viewers have no idea what all this is about. FACT.
Click to expand...
I've never understood what the "Thread Association Game" thread is all about, but that's got over 29,000 posts on it, so I leave it for those who enjoy it.

Given most of these stats are derived from "xG" figures, would you like an explanation of it?
 
DHCanary said:
I've never understood what the "Thread Association Game" thread is all about, but that's got over 29,000 posts on it, so I leave it for those who enjoy it.

Given most of these stats are derived from "xG" figures, would you like an explanation of it?
Click to expand...
Yes please
 
DHCanary said:
I've never understood what the "Thread Association Game" thread is all about, but that's got over 29,000 posts on it, so I leave it for those who enjoy it.
Click to expand...
I'll explain it if you're foxed.
Nice of you to leave it up for us of limited intellect. <ok>
 
  • Like
Reactions: Bure budgie
Bure budgie said:
Yes please
Click to expand...
xG is Expected Goals.
Essentially, it measures the quality of a chance, and how often a similar chance is converted by the "average" striker. That factors in the location on the pitch, how the final pass was made, whether its a shot or header, etc. Some more advanced models also factor in the position of other players on the pitch, whether it's the strikers dominant foot, etc. By comparing a particular chance to a massive database of previous shots taken, an xG value between 0 and 1 is assigned to a shot. So an xG of 0.1 means that a striker would normally take that opportunity 10% of the time. An xG of 0.9 would mean the striker would normally score 90% of the time - so a much better chance.

That xG value assigned to a shot can then be grouped with the xG of other shots, to measure the performance of a particular team/player. So if all the shots a player takes in a game add up to an xG of 0.2, but he scores 2 goals, then he's done much better than expected. Generally, that means he's either an elite finisher (Messi, Ronaldo, etc can consistently outperform xG), or got a bit lucky and will be unlikely to keep scoring that much. Similarly, Patrick Bamford consistently under-performs xG. He might come out of a game with an xG of 2.3, and score 0 or 1 goals. Bamford gets into good positions and Leeds create loads of chances, so he scores a good number despite not being excellent at finishing. Last season in the Championship, Bamford scored 16 goals in the League, but the chances he had were assigned an xG value of 35. Whilst Pukki in 18/19 had an xG of 22.7, but scored 29 goals. Pukki is an excellent finisher, given the chances Bamford had last season, Pukki at Leeds could easily have scored 40 goals.

Similarly for a team, you can sum up the xG of all their players across a game/season, and assess how good they are at creating/taking chances. So you can compare clubs to each other by comparing xG values. In particular, comparing xG to actual goals gives an indication of how "lucky" the team have been (An xG of 10 but only scoring 1 goal indicates dreadful luck, or a team with woeful finishers), and can also be used to infer how sustainable performances are. If a team are scoring more goals than they've created the chances to justify, you might expect them to struggle to maintain their form in the league.

xGA is Expected Goals Against. If a shot is taken against a defence with an xG value of 0.9, then that defence have given up a very good chance. Over a game, if a team have an xGA of 2.5, but keep a clean sheet, they haven't really defended that well and have been bailed out by poor finishing or an excellent goalkeeping performance. So you can assess over a longer period of time whether a defence are good at keeping an opposition quiet. Looking at more detail at this statistic can give other insights or reflections on the way a club play. The defensive policy of Burnley for example is to make sure they give absolutely nothing away in their own penalty box. They defend set pieces well, crosses into the box, etc. Over a game, the xG for shots in the penalty area is usually very low. However, Burnley are less worried about shots from distance, and don't prevent them to the same degree as other teams. So most of their xGA is from conceding lower risk shots from distance.

Looking at the analysis of goalkeepers above, the x axis for "xG performance" is measuring how many fewer goals a goalkeeper has conceded, compared to how many would be expected from the chances they've faced. Krul is top of this metric, he's conceded 0.8 fewer goals than would be expeceted from the quality of shots he's faced. At the opposite end of the scale, Cabral and Rudd are performing worse than would be expected. McGovern is buried in the middle of that chart, performing slightly better than average.
 
  • Like
Reactions: Bure budgie and robbieBB
DHCanary said:
xG is Expected Goals.
Essentially, it measures the quality of a chance, and how often a similar chance is converted by the "average" striker. That factors in the location on the pitch, how the final pass was made, whether its a shot or header, etc. Some more advanced models also factor in the position of other players on the pitch, whether it's the strikers dominant foot, etc. By comparing a particular chance to a massive database of previous shots taken, an xG value between 0 and 1 is assigned to a shot. So an xG of 0.1 means that a striker would normally take that opportunity 10% of the time. An xG of 0.9 would mean the striker would normally score 90% of the time - so a much better chance.

That xG value assigned to a shot can then be grouped with the xG of other shots, to measure the performance of a particular team/player. So if all the shots a player takes in a game add up to an xG of 0.2, but he scores 2 goals, then he's done much better than expected. Generally, that means he's either an elite finisher (Messi, Ronaldo, etc can consistently outperform xG), or got a bit lucky and will be unlikely to keep scoring that much. Similarly, Patrick Bamford consistently under-performs xG. He might come out of a game with an xG of 2.3, and score 0 or 1 goals. Bamford gets into good positions and Leeds create loads of chances, so he scores a good number despite not being excellent at finishing. Last season in the Championship, Bamford scored 16 goals in the League, but the chances he had were assigned an xG value of 35. Whilst Pukki in 18/19 had an xG of 22.7, but scored 29 goals. Pukki is an excellent finisher, given the chances Bamford had last season, Pukki at Leeds could easily have scored 40 goals.

Similarly for a team, you can sum up the xG of all their players across a game/season, and assess how good they are at creating/taking chances. So you can compare clubs to each other by comparing xG values. In particular, comparing xG to actual goals gives an indication of how "lucky" the team have been (An xG of 10 but only scoring 1 goal indicates dreadful luck, or a team with woeful finishers), and can also be used to infer how sustainable performances are. If a team are scoring more goals than they've created the chances to justify, you might expect them to struggle to maintain their form in the league.

xGA is Expected Goals Against. If a shot is taken against a defence with an xG value of 0.9, then that defence have given up a very good chance. Over a game, if a team have an xGA of 2.5, but keep a clean sheet, they haven't really defended that well and have been bailed out by poor finishing or an excellent goalkeeping performance. So you can assess over a longer period of time whether a defence are good at keeping an opposition quiet. Looking at more detail at this statistic can give other insights or reflections on the way a club play. The defensive policy of Burnley for example is to make sure they give absolutely nothing away in their own penalty box. They defend set pieces well, crosses into the box, etc. Over a game, the xG for shots in the penalty area is usually very low. However, Burnley are less worried about shots from distance, and don't prevent them to the same degree as other teams. So most of their xGA is from conceding lower risk shots from distance.

Looking at the analysis of goalkeepers above, the x axis for "xG performance" is measuring how many fewer goals a goalkeeper has conceded, compared to how many would be expected from the chances they've faced. Krul is top of this metric, he's conceded 0.8 fewer goals than would be expeceted from the quality of shots he's faced. At the opposite end of the scale, Cabral and Rudd are performing worse than would be expected. McGovern is buried in the middle of that chart, performing slightly better than average.
Click to expand...
Thanks for that DH.
Although I must admit that most of it goes over my head,I do appreciate that a lot of people enjoy the stats.
Seasons greetings to one and all.<cracker><cracker>
 
  • Like
Reactions: DUNCAN DONUTS, DHCanary and robbieBB
DHCanary said:
xG is Expected Goals.
Essentially, it measures the quality of a chance, and how often a similar chance is converted by the "average" striker. That factors in the location on the pitch, how the final pass was made, whether its a shot or header, etc. Some more advanced models also factor in the position of other players on the pitch, whether it's the strikers dominant foot, etc. By comparing a particular chance to a massive database of previous shots taken, an xG value between 0 and 1 is assigned to a shot. So an xG of 0.1 means that a striker would normally take that opportunity 10% of the time. An xG of 0.9 would mean the striker would normally score 90% of the time - so a much better chance.

That xG value assigned to a shot can then be grouped with the xG of other shots, to measure the performance of a particular team/player. So if all the shots a player takes in a game add up to an xG of 0.2, but he scores 2 goals, then he's done much better than expected. Generally, that means he's either an elite finisher (Messi, Ronaldo, etc can consistently outperform xG), or got a bit lucky and will be unlikely to keep scoring that much. Similarly, Patrick Bamford consistently under-performs xG. He might come out of a game with an xG of 2.3, and score 0 or 1 goals. Bamford gets into good positions and Leeds create loads of chances, so he scores a good number despite not being excellent at finishing. Last season in the Championship, Bamford scored 16 goals in the League, but the chances he had were assigned an xG value of 35. Whilst Pukki in 18/19 had an xG of 22.7, but scored 29 goals. Pukki is an excellent finisher, given the chances Bamford had last season, Pukki at Leeds could easily have scored 40 goals.

Similarly for a team, you can sum up the xG of all their players across a game/season, and assess how good they are at creating/taking chances. So you can compare clubs to each other by comparing xG values. In particular, comparing xG to actual goals gives an indication of how "lucky" the team have been (An xG of 10 but only scoring 1 goal indicates dreadful luck, or a team with woeful finishers), and can also be used to infer how sustainable performances are. If a team are scoring more goals than they've created the chances to justify, you might expect them to struggle to maintain their form in the league.

xGA is Expected Goals Against. If a shot is taken against a defence with an xG value of 0.9, then that defence have given up a very good chance. Over a game, if a team have an xGA of 2.5, but keep a clean sheet, they haven't really defended that well and have been bailed out by poor finishing or an excellent goalkeeping performance. So you can assess over a longer period of time whether a defence are good at keeping an opposition quiet. Looking at more detail at this statistic can give other insights or reflections on the way a club play. The defensive policy of Burnley for example is to make sure they give absolutely nothing away in their own penalty box. They defend set pieces well, crosses into the box, etc. Over a game, the xG for shots in the penalty area is usually very low. However, Burnley are less worried about shots from distance, and don't prevent them to the same degree as other teams. So most of their xGA is from conceding lower risk shots from distance.

Looking at the analysis of goalkeepers above, the x axis for "xG performance" is measuring how many fewer goals a goalkeeper has conceded, compared to how many would be expected from the chances they've faced. Krul is top of this metric, he's conceded 0.8 fewer goals than would be expeceted from the quality of shots he's faced. At the opposite end of the scale, Cabral and Rudd are performing worse than would be expected. McGovern is buried in the middle of that chart, performing slightly better than average.
Click to expand...
Thanks from me as well DH. That's a very clear explanation.
 
Only stat that counts Top of the league 22 games played 44 points accumulated.

Average of 2 points per game. Despite the worse injury crises the team has faced since DF was appointed.
 
Last edited:
If 2020 was a league table:
You must log in or register to see media

Bit arbitrary, but at least highlights how crazy Brentford's form has been this year. 23 wins in 43 games. Also Stevenage with 4 wins in a calendar year, ouch.
 
You must log in or register to reply here.
Top Bottom
Back