By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions. We in turn value your personal details in accordance with our Privacy Policy.
Please log in or register. Registered visitors get fewer ads.
It seems to conclude that we're pretty efficient outfit so far, but I must say I did chuckle at this statement: "One of the main reasons for their success is the number of goals scored by the team." Yeah, that's how football kind of works!
1
Apparently, we have low xG on 17:02 - Sep 3 with 1025 views
Apparently, we have low xG on 16:41 - Sep 3 by Northernr
Trying to explain my problem, other than as Rsonist says the sanctimonious delivery of it by people who believe in it...
As I understand it (please correct this if it's wrong) xG takes all the efforts on goal across the division and uses them to calculate the percentage number of headers from 8 yards that go in, or shots from 15 yards, or volleys from 30 yards, and then gives every action a score between 0 and 1 of it's likelihood of a goal. So a right footed shot from six yards out centre of goal goes in 80% of the time, 0.8, and a 30 yarder goes in 1% of the time, so 0.01. QPR could win 3-0 with three shots on target from 30 yards and have an xG of 0.03, or lose 2-0 having missed an absolute sitter from six yards and have an xG of 0.8.
What it doesn't take any account of, again as I understand it, is who's taking the shot. QPR's xG at Reading next Saturday will be the same with Austin, Dykes and Gray up front as it would have been if we were still picking Washington, Sylla and Polter. This is the flaw in it for me, you can't just point and say "QPR will definitely fall back, they can't possibly keep this up" because maybe we've just got good finishers, and unless they get injured we will keep this up.
If you could give me a stat that took the first paragraph and corrected the second. Combining the divisional average of shots that go in from 15 yards, with the average number of Charlie Austin shots that go in from 15 yards, and marry those two together, then I think you get a much better number that I'd be right on board with. For me you can't say "well it's a header from ten yards so it's only 0.2 chance of a goal" if the guy taking the header is Les fcking Ferdinand.
Enjoying the thread though.
This post has been edited by an administrator
"Building on xG, we can dig deeper into the likelihood of a shot becoming a goal. Whereas xG provides a value of the shot before the player shoots, expected goals on target (xGOT) provides a modified value of an on-target shot after the player shoots.
The value presented is the same – a number between zero (no chance of a goal) and one (a certain goal). But it adds further context by crediting the shots that head towards the top corner vs those that are straight down the middle of the goal. As you would imagine, this xGOT model only calculates a value for shots that are on target, so relies on the player to at least work the keeper to be registered – of course, if the shot is off-target your chance of scoring a goal is zero (barring an unlikely deflection).
Let’s take this example below from Harry Kane last season. Before he takes the shot outside the area, his xG is 0.03 – a low-value chance that would be expected to be a goal only three times out of 100.
However, the quality of the strike from Kane is outstanding, as it flies into the top corner. Having taken the shot, the xGOT value of this chance now jumps up to 0.54, meaning that the location of that strike would lead to a goal 54 per cent of the time. A high-quality finish indeed.
The main thing xGOT provides is a clearer idea of a player’s shot execution. If a player has an xGOT value that is consistently higher than their xG, this tells us that they are shooting at a better rate than the quality of the chances they are getting.
It is worth noting that such figures do only include shots that hit the target, which unfortunately excludes blocked shots. Therefore, there may be some players who can count themselves unlucky that they executed the perfect shot but the defender simply got in the way.
We can then use Opta’s metric of “shooting goals added” which calculates the difference between a player’s xG and xGOT. Below, you can see that Tottenham Hotspur’s Son Heung-min added the most value to his shots last season, by improving his chances of scoring by 3.8 goals from the placement of those shots."
Again from worville on the athletic. Hope this helps.
1
Apparently, we have low xG on 17:23 - Sep 3 with 982 views
Apparently, we have low xG on 17:02 - Sep 3 by slmrstid
I'm not sure I remember Brett Angell ever having a shot, let alone from 12 yards.
Actually, considering the weight he was when he played for us I'd be surprised if he ever managed to get close enough to the penalty box...!
How very dare you he once hit the bar from the edge of the box and whilst he was playing for us if VAR he might be offside at kickoff just from his belly being in the penalty area
0
Apparently, we have low xG on 17:26 - Sep 3 with 975 views
Apparently, we have low xG on 16:59 - Sep 3 by Antti_Heinola
hmmm... other than the factual inaccuracies here (Barbet was under pressure, no keeper on earth would've held that shot, Austin got there first not by 'luck' but because of great reactions and instinct etc etc) what you are describing is how a team ball game tends to work - split second decisions, bounces of the ball, all the many, many marginal gains that happen every second. To that degree everything that happens in a match is 'luck'.
I'm similar to norf here in that I don't see the harm in the stats, as with everything, they are a useful tool in a much larger toolbox. To some degree, the stat with us is correct. We're unbeaten, but actually several games *could* have gone the other way, although you didn't necessarily need XG to tell you that. Even Warbs said we should have lost to Orient, and should've been losing at HT to Cov.
The clickbaity tweet is also definitely right. We won't keep this up - ie, we will not go the season unbeaten. We are going to lose games. But our form since Jan is a much better guide than a very small sample size of matches.
I'm going to ignore your first point, as i caveated Austin and the keeper's roles in the luck of this goal in my original post, and the first is, lets say, arguable. There is a book called the numbers game, where a whole chapter discusses evidence that football is a 50-50 sport (chance-skill). Luck is certainly involved throughout a football match, and in every action. Obviously agree with the last two points as I said them in the original post
0
Apparently, we have low xG on 17:35 - Sep 3 with 962 views
Apparently, we have low xG on 16:55 - Sep 3 by Northernr
Yeh. Maybe that will happen. It's every bit as likely as us falling back inline with the xG. As you can tell, it's this high and mighty "oh QPR can't possibly keep this up" that's irked me.
Had a look at the x G site and strangely enough there are no stats included for Leicester and their 2015/16 Title winning season when their tactic of countering teams on the break and being clinical would not score well regards xG. What it will tell you is if a team is creating chances and aren't being clinical enough then time to get yourself better forwards
0
Apparently, we have low xG on 17:35 - Sep 3 with 962 views
Expected Goals (xG) is a statistic that can be used in a number of ways.
From an attacking point of view, if the xG score for a team is low it means they are not creating as many chances to score as a team with a high xG. It is better than simply a shots on/off target measure as it takes into account a measure of the quality of the chance. So what should this tell the manager? It tells him it’s not necessarily a problem with the forwards but with the build-up play.
It doesn’t take into account who is taking the shot, but this isn’t necessarily a flaw as it allows it to be used to tell us things about the player that shoots. Firstly, if his scoring rate is higher than the xG for his shots then he is a better than average shot taker, this can be analysed not only for overall shot taking ability but can be analysed on a position by position basis i.e. good from close in on the left, good from distance, poor with his head etc.
It’s also not all about the man who takes the shot either, xG can be tracked back to the build-up play i.e. a winger can be getting in brilliant crosses, it’s not his fault the forward keeps missing. This can help in scouting i.e. we know we are creating certain types of chances (the xG for these are high) but our forwards are worse than average on these types of chances (putting in less goals then the xG suggests) therefore we should look for players who have a better than average return for these types of chances. On the other hand we might have a striker who bangs in more than average of a certain type of goal, but the midfield don’t create to many of these chances, therefore we should scout for a player with a history of creating just those types of chances. Also if you not only look at your own xG, but look at your opponents you can see how you’re your defence is at reducing an opponent’s chances i.e. opponents who normally play in a certain style e.g. a high press, normally average 2.1 xG per game but we reduce them to a 1.8, on the other hand opponents who sit back and play on the break average say 1.9 xG but against us average 2.2, this suggests we need to prioritise practicing our defence against this type of play.
So what would we rather be, a team like we have now, with a lower that average xG but are clinical in putting what we do create away (higher goals than xG), or a team that creates loads of quality chances (high xG) but fails to put many away (lower goals than xG)? Both could give exactly the same results as we have now. Or just maybe can we dream of a team that both creates chances AND puts them away, and maybe it’s by comparing our xG to actual goals that lets the manager know what needs work.
1
Apparently, we have low xG on 18:01 - Sep 3 with 942 views
Apparently, we have low xG on 16:41 - Sep 3 by Northernr
Trying to explain my problem, other than as Rsonist says the sanctimonious delivery of it by people who believe in it...
As I understand it (please correct this if it's wrong) xG takes all the efforts on goal across the division and uses them to calculate the percentage number of headers from 8 yards that go in, or shots from 15 yards, or volleys from 30 yards, and then gives every action a score between 0 and 1 of it's likelihood of a goal. So a right footed shot from six yards out centre of goal goes in 80% of the time, 0.8, and a 30 yarder goes in 1% of the time, so 0.01. QPR could win 3-0 with three shots on target from 30 yards and have an xG of 0.03, or lose 2-0 having missed an absolute sitter from six yards and have an xG of 0.8.
What it doesn't take any account of, again as I understand it, is who's taking the shot. QPR's xG at Reading next Saturday will be the same with Austin, Dykes and Gray up front as it would have been if we were still picking Washington, Sylla and Polter. This is the flaw in it for me, you can't just point and say "QPR will definitely fall back, they can't possibly keep this up" because maybe we've just got good finishers, and unless they get injured we will keep this up.
If you could give me a stat that took the first paragraph and corrected the second. Combining the divisional average of shots that go in from 15 yards, with the average number of Charlie Austin shots that go in from 15 yards, and marry those two together, then I think you get a much better number that I'd be right on board with. For me you can't say "well it's a header from ten yards so it's only 0.2 chance of a goal" if the guy taking the header is Les fcking Ferdinand.
Enjoying the thread though.
This post has been edited by an administrator
xG is how good we are at creating chances Goals above/below xG is how well we are doing at converting them
As far as I can see your problem is less with xG itself than the way it's interpreted and used by people. It's not meant to be a metric to say how lucky we are (which is how it's being used in the suggestion that we'll revert to norm) it's a metric to say how well we are creating chances, and therefore goals above xG is how well we are doing with converting them
0
Apparently, we have low xG on 23:46 - Sep 3 with 788 views
Apparently, we have low xG on 16:32 - Sep 3 by Northernr
Again, as far as I understand it, xG basically counts shots like Barbet's from 40 yards as like 0.1 chance of a goal, so it would go down as luck in that theory that we got a goal off it. Barbet's follow up shot, again as I understand it, scored quite highly. Where this falls down is if you have somebody like Dickie, or Adam Reach, who is very good from that range, and it's not luck at all it's a skill like any other, but xG doesn't care if it's Reach, Messi, Barbet or Dominic Iorfa shooting from 30 yards, they all count the same. Which is a flaw. IMO.
I see, understand, but like you don't agree!
So Waghorn gets more 'credit' because he missed from inside the box - very good Seny save, than Barbet does for a peach of a shot from 30+ yards? - Bonkers!
0
Apparently, we have low xG on 10:49 - Sep 4 with 692 views
Apparently, we have low xG on 16:32 - Sep 3 by mobileDeWijs
With all due respect Paul, there's quite a lot of luck in the sequence you described. The CB Barbet being in a shooting position, under no pressure, the failure by the 2 cov defenders to first win the parry and then block austin, and crucially, the '[even] better finish' from Barbet. Whilst all these parts individually could happen (great low, hard shot from Barbet, Austin's positioning, Barbet continuing his run into the box, and, most glaringly, the repeated failure by coventry to win a second ball, including from the header away from Johanse's inital cross, plus the failure by the keeper to hold onto the ball, admittedly difficult though), it would be harsh to say there wasnt some fortune in the sequence which led to the goal. Unless we practiced that one on the training ground, defensive errors and all! Im not complaing though, great moment and they all count.
sorry, that's not luck but our good play and their bad.
By your criteria, every goal, has an element of 'luck'.
Every goal is someone's fault and 'could' have been prevented.
However if you do things correctly and the opposition don't then that's not luck!
As someone said above, missed handball by the ref, deflected shot that wasn't going in, defender falls over - all examples of luck. I'm not having that most people would look at that goal and say "that was lucky"
0
Apparently, we have low xG on 12:02 - Sep 4 with 655 views
Apparently, we have low xG on 10:49 - Sep 4 by PinnerPaul
sorry, that's not luck but our good play and their bad.
By your criteria, every goal, has an element of 'luck'.
Every goal is someone's fault and 'could' have been prevented.
However if you do things correctly and the opposition don't then that's not luck!
As someone said above, missed handball by the ref, deflected shot that wasn't going in, defender falls over - all examples of luck. I'm not having that most people would look at that goal and say "that was lucky"
I would argue that in every goal - in every action in a football match - 'luck' is at least partially if not significantly involved, as ultimately, no matter our or the opponents skill, every action cannot be completely controlled.
I suppose we just see that goal differently though
0
Apparently, we have low xG on 12:03 - Sep 4 with 653 views
Apparently, we have low xG on 10:49 - Sep 4 by PinnerPaul
sorry, that's not luck but our good play and their bad.
By your criteria, every goal, has an element of 'luck'.
Every goal is someone's fault and 'could' have been prevented.
However if you do things correctly and the opposition don't then that's not luck!
As someone said above, missed handball by the ref, deflected shot that wasn't going in, defender falls over - all examples of luck. I'm not having that most people would look at that goal and say "that was lucky"
I’d say there was an element of luck in how the ball got from Austin to Barbet. It wasn’t intentional, it was a shot I think but that bit of luck was fully deserved after the initial shot and run into the box. Great determination from Barbet.
Out of interest would Austin’s mishit shot score a higher xg score than Barbets piledriver?
Apparently, we have low xG on 12:02 - Sep 4 by mobileDeWijs
I would argue that in every goal - in every action in a football match - 'luck' is at least partially if not significantly involved, as ultimately, no matter our or the opponents skill, every action cannot be completely controlled.
I suppose we just see that goal differently though
Agree with the last line!
0
Apparently, we have low xG on 13:20 - Sep 4 with 596 views
In my 20's I found out I had a low Xg, apparently it was my fondness for very hot showers. All OK now and I am proud to say that I am a grandfather to two lovely little girls. Moral of the story: a low Xg doesn't mean your any less of a man.
0
Apparently, we have low xG on 14:13 - Sep 4 with 564 views
Apparently, we have low xG on 16:41 - Sep 3 by Northernr
Trying to explain my problem, other than as Rsonist says the sanctimonious delivery of it by people who believe in it...
As I understand it (please correct this if it's wrong) xG takes all the efforts on goal across the division and uses them to calculate the percentage number of headers from 8 yards that go in, or shots from 15 yards, or volleys from 30 yards, and then gives every action a score between 0 and 1 of it's likelihood of a goal. So a right footed shot from six yards out centre of goal goes in 80% of the time, 0.8, and a 30 yarder goes in 1% of the time, so 0.01. QPR could win 3-0 with three shots on target from 30 yards and have an xG of 0.03, or lose 2-0 having missed an absolute sitter from six yards and have an xG of 0.8.
What it doesn't take any account of, again as I understand it, is who's taking the shot. QPR's xG at Reading next Saturday will be the same with Austin, Dykes and Gray up front as it would have been if we were still picking Washington, Sylla and Polter. This is the flaw in it for me, you can't just point and say "QPR will definitely fall back, they can't possibly keep this up" because maybe we've just got good finishers, and unless they get injured we will keep this up.
If you could give me a stat that took the first paragraph and corrected the second. Combining the divisional average of shots that go in from 15 yards, with the average number of Charlie Austin shots that go in from 15 yards, and marry those two together, then I think you get a much better number that I'd be right on board with. For me you can't say "well it's a header from ten yards so it's only 0.2 chance of a goal" if the guy taking the header is Les fcking Ferdinand.
By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions.
