Jump to content
Sign in to follow this  
hogesar

Expected Goals (Points) Table Updated

Recommended Posts

Expected goals = bulls**t. Smash one into the net from 30 yards: 1 actual goal and 0.1 expected goals; smash one from 35 yards, keeper makes a great save but provides striker with an open goal from 10 yards out who nets: 1 actual goal and 0.8 expected goals….ditto penalties. 

Share this post


Link to post
Share on other sites
25 minutes ago, Google Bot said:

How come Brentford were expected to be so high?  Someone explain where this comes from?

Its based on "Expected goals" for and against.

 

https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.thepunterspage.com/expected-goals-explained/&ved=2ahUKEwiC77ehsbTzAhUnQkEAHSgADNsQFnoECB4QAQ&usg=AOvVaw0ayZUbOqNCKpP45wPyuv3k

Share this post


Link to post
Share on other sites
4 hours ago, Jezzard said:

Expected goals = bulls**t. Smash one into the net from 30 yards: 1 actual goal and 0.1 expected goals; smash one from 35 yards, keeper makes a great save but provides striker with an open goal from 10 yards out who nets: 1 actual goal and 0.8 expected goals….ditto penalties. 

Well that doesn't make it sh*t, that's literally what it does. The chances of scoring from 35 yards are far less than scoring a penalty, shockingly.

Share this post


Link to post
Share on other sites

10 hours ago, hogesar said:

Its based on "Expected goals" for and against.

I don't get why Brentford have such a higher expected goals ratio versus us though?  Last season we scored similar to them yet picked up more points.  

Or was this table drawn up after a few matches of the season perhaps?  I'm presuming it was a pre-season prediction so maybe that's where I'm confused.

I guess what I'm asking is why so many people thought Brentford would do well, and they've so far been proven right.  And so many people thought we'd fail, and so far that's looking our path too.  I get that subjectively there's this 'new team' mentality that they carried into the prem, but objectively the stats appear to back it up too.

Feels like we're missing out on a piece of the pie when the stats back it up.

Share this post


Link to post
Share on other sites
15 minutes ago, Google Bot said:

I don't get why Brentford have such a higher expected goals ratio versus us though?  Last season we scored similar to them yet picked up more points.  

Or was this table drawn up after a few matches of the season perhaps?  I'm presuming it was a pre-season prediction so maybe that's where I'm confused.

I guess what I'm asking is why so many people thought Brentford would do well, and they've so far been proven right.  And so many people thought we'd fail, and so far that's looking our path too.  I get that subjectively there's this 'new team' mentality that they carried into the prem, but objectively the stats appear to back it up too.

Feels like we're missing out on a piece of the pie when the stats back it up.

Its based on the quality of goal scoring chances created in matches so far this season. Its saying that based on the chances we've had we would have beenb expected to score 5 goals but we've actually scored 1. Conclusion being we are not clinical enough.

So Brentford have obviously created more good goalscoring chances than us so far this season.

Sometimes teams can buck the trend for a whole season but usually if you are much higher in the second column than the first its a sign that your form/position may be a little false because it means as a team you are being unbelievably clinical which is unlilely to be sustained over a full season. Interestingly Brighton (our next opponants) are one team you would put in that category at present.  

Share this post


Link to post
Share on other sites

This really does my head in.

So if a team does better than their xg suggests, it's kind of false, if they are doing worse, it's because they aren't clinical enough - or in layman's terms, they are on form or they are off form.  If they are performing as xgs suggest, then that is seen as some kind of norm and validification of the xgs. Wow.

Honestly, the more I read about this xg stuff, the more I think.....just watch the matches and use your eyes and brain. There are SO many variables when it comes to these calculations that it is becoming quite bizarre that this has caught on.  So they can look of interest on the surface, but the more you delve into it, the more questions that need to be asked to make allowance for all the different variables.  All you get is a never ending and increasing number of parameters that have to be taken into account, that it becomes impossible to take them seriously. The more you dissect something, the further away you get away from the whole - and all you get is more questions.

If we're not careful, football will start looking like a computer game with players heads so full of statistical nonsense that they are incapable to think for themselves. 

 

Share this post


Link to post
Share on other sites
31 minutes ago, Jim Smith said:

Its saying that based on the chances we've had we would have beenb expected to score 5 goals but we've actually scored 1. Conclusion being we are not clinical enough.

But the chart says points, not goals(?).   We've scored 2.  And Brentford haven't score 12 but they have 12 points.

I still don't get my head around why Brentford statistically were predicted to be 13 points good within 7 games, yet teams like Arsenal and Spurs were expected to be at 7-8 points.  There just seems to be no sense there, and if they're using the championship as a gauge we should be around Brentford in the prediction as goal tallies were similar.

Because the stats, when related to us and Brentford are spot on in regards to league positions, so something is right in this prediction... Question is, why couldn't we see it, and could we have used it to make key signings?! 

Or perhaps it's that removing Emi from being a big portion of that goal creation has sunk us down perhaps??  Whatever was used, has been proven correct, regardless.

Edited by Google Bot

Share this post


Link to post
Share on other sites
8 minutes ago, lake district canary said:

This really does my head in.

So if a team does better than their xg suggests, it's kind of false, if they are doing worse, it's because they aren't clinical enough - or in layman's terms, they are on form or they are off form.  If they are performing as xgs suggest, then that is seen as some kind of norm and validification of the xgs. Wow.

Honestly, the more I read about this xg stuff, the more I think.....just watch the matches and use your eyes and brain. There are SO many variables when it comes to these calculations that it is becoming quite bizarre that this has caught on.  So they can look of interest on the surface, but the more you delve into it, the more questions that need to be asked to make allowance for all the different variables.  All you get is a never ending and increasing number of parameters that have to be taken into account, that it becomes impossible to take them seriously. The more you dissect something, the further away you get away from the whole - and all you get is more questions.

If we're not careful, football will start looking like a computer game with players heads so full of statistical nonsense that they are incapable to think for themselves. 

 

Its not false, your results are what they are but if there is a big discrepancy between Xg and actual goals (whether attacking or defending) then it can be an indicator that your form is not sustainable long term.

Not necessarily though, it could equally mean you have a really good striker bang in form or a great goalkeeper or indeed that you are just lucky/unlucky. 

Swansea were a good example of it last season. They were right up the top for a long period despite the fact that Xg showed they were scoring a very high percentage of chances and conceding less goals than would be expected. They fell away eventually and whilst they remained in playoff contention they never really threatened the top 2 again. 

  • Like 1

Share this post


Link to post
Share on other sites
2 minutes ago, Google Bot said:

But the chart says points, not goals(?).   We've scored 2.  And Brentford haven't score 12 but they have 12 points.

I still don't get my head around why Brentford statistically were predicted to be 13 points good within 7 games, yet teams like Arsenal and Spurs were expected to be at 7-8 points.  There just seems to be no sense there, and if they're using the championship as a gauge we should be around Brentford in the prediction as goal tallies were similar.

Because the stats, when related to us and Brentford are spot on in regards to league positions, so something is right in this prediction... Question is, why couldn't we see it, and could we have used it to make key signings?! 

Or perhaps it's that removing Emi from being a big portion of that goal creation has sunk us down perhaps??  Whatever was used, has been proven correct, regardless.

Sorry I miss-read that.

Same point applies - its saying if you took our results based on expected goals then you would expect us to have 5 points so you could argue it shows we've been a bit unlucky and not clinical enough. 

Share this post


Link to post
Share on other sites
4 minutes ago, Google Bot said:

But the chart says points, not goals(?).   We've scored 2.  And Brentford haven't score 12 but they have 12 points.

I still don't get my head around why Brentford statistically were predicted to be 13 points good within 7 games, yet teams like Arsenal and Spurs were expected to be at 7-8 points.  There just seems to be no sense there, and if they're using the championship as a gauge we should be around Brentford in the prediction as goal tallies were similar.

Because the stats, when related to us and Brentford are spot on in regards to league positions, so something is right in this prediction... Question is, why couldn't we see it, and could we have used it to make key signings?! 

Or perhaps it's that removing Emi from being a big portion of that goal creation has sunk us down perhaps??  Whatever was used, has been proven correct, regardless.

Its not a prediction. Its looking at the quality of chances created in the games so far and what the result of those games would ordinarily be expected to be based on that. I don;t have the stats to hand but I suspect that based on Xg the Burnley game would be expected to be a draw, probably Everton and maybe Leicester we might have been expected to win? I'm not 100% sure though. 

  • Like 1

Share this post


Link to post
Share on other sites

10 minutes ago, Jim Smith said:

Its not a prediction. Its looking at the quality of chances created in the games so far and what the result of those games would ordinarily be expected to be based on that.

Ahhhh ok, so that's making sense now.  Ok.  So it's more a display that we're under performing vs our expected potential in each game so far?  The penny has dropped as Brentford have been creating more chances and also converting them at their expected rate, whereas we've been not only creating less chances but also failing at taking them - thank you! 🙂

The 'expected' chart on the left is more akin to the chances we've created, and the chart on the right is how we've preformed vs those chances, effectively.

Edited by Google Bot
  • Thanks 1

Share this post


Link to post
Share on other sites
1 minute ago, Google Bot said:

Ahhhh ok, so that's making sense now.  Ok.  So it's more a display that we're under performing vs our expected potential in each game so far?  The penny has dropped as Brentford have been creating more chances and also converting them at their expected rate then  - thank you! 🙂

Yes - also factors in chances given to the opposition as well I think. So Bretford may also be high because they are tight defensively ad don;t allow the opposition too many good chances.

I suspect defensively we are not quite as bad as would be expected because in several of our games the opposition have scored from more or less their only decent chances. Arsenal aside (maybe Watford too) its not like they've had endless good chances. Man City actually scored 5 from 4 shots as I recall!

Share this post


Link to post
Share on other sites
22 hours ago, hogesar said:

Well that doesn't make it sh*t, that's literally what it does. The chances of scoring from 35 yards are far less than scoring a penalty, shockingly.

Not what I wrote. In the first scenario, the keeper making the save increased probability of scoring/  XG from 0.1 to 0.9 without the attacking team doing anything. In the penalty scenario, the XG at the point when a player is fouled is variable and usually much lower than the 0.85 or so XG for the penalty kick - see Lees-Melou v Leicester for example;  when he was fouled, the XG was about 0.05. These are are but two examples where XG is biased (sh1t). Another more common one is when a team is in the lead tactically sits back and allows the opponent to have numerous half chances (0.1 XG) which accumulate and may then exceed that of the winning team after 90 minutes. XG can be useful as long as its limitations is acknowledged, but is interpreted by many pundits as forming a justice league, for which it is inadequate.       

Share this post


Link to post
Share on other sites

It is pretty adequate at showing who is the most clinical, conversion rates say a lot about the quality of a team overall. It’s interesting to see Brighton outperforming their XG. 
 

one thing  I do feel is this does Pukki a disservice as he’s been fed scraps so far this season I’ve no doubt if he was given more chances he would convert more freely than some strikers in the league. I bet if he’s given 3 clear chances he’ll score at least one a game. 

Share this post


Link to post
Share on other sites

Has anyone seen Experimental365's prem graphs? They had to extend the size of the graph to plot our attacks ineffectiveness....

 

FBFCS0qWUAYtus-.png

Share this post


Link to post
Share on other sites
29 minutes ago, hogesar said:

Has anyone seen Experimental365's prem graphs? They had to extend the size of the graph to plot our attacks ineffectiveness....

 

FBFCS0qWUAYtus-.png

Wow

Share this post


Link to post
Share on other sites
53 minutes ago, Jim Smith said:

Wow

Yeah, uhhh, I hadn't actually realised till seeing the graph. Slight cause for concern.. I suppose it's purely because we've not scored any goals. Score a couple against Brighton and we're probably in line with the likes of Arsenal / Wolves.

Share this post


Link to post
Share on other sites

2 hours ago, hogesar said:

Has anyone seen Experimental365's prem graphs? They had to extend the size of the graph to plot our attacks ineffectiveness....

Dividing 70+ shots into 2 goals gives a shocking result lol, if we got the decision against Leicester and Pukki wasn't pulled back vs Burnley and scored it would look a whole lot rosier and we'd be (negatively) higher than spurs.

Unfortunately, none of those things happened and we're off the standard charts. 😞

Share this post


Link to post
Share on other sites
57 minutes ago, Duncan Edwards said:

Quick!! Send this to Talksport!!

 

Yeah, like they'd understand a graph.

  • Haha 1

Share this post


Link to post
Share on other sites
2 hours ago, hogesar said:

Has anyone seen Experimental365's prem graphs? They had to extend the size of the graph to plot our attacks ineffectiveness....

 

FBFCS0qWUAYtus-.png

That is quite something! But always going to look like that when you have scored virtually nothing

Share this post


Link to post
Share on other sites

xG is not perfect but it's more useful than most of the stats we get e.g. number of corners?  Possession %? These tell you nothing about who played better, they just tell you who had more corners or more possession.

 

xG is a recognition that football is often a game of fine margins, and from any given chance there is a % likelihood whether or not you score, sometimes they go in and sometimes they don't.  The actual scoreline in football can be quite misleading because e.g. a team can win 2-0 because they are fortunate with a couple of low % chances being scored despite the opposition having many better chances but not putting any of them away.  For managers, they will try to setup their team so that they create good scoring chances while minimising those of the opposition, but they can't guarantee their attackers will actually take those chances.

 

There is certainly a lazy response to xG stats which shows a lack of understanding, which is that if your xG is higher than actual goals scored, you are "not being clinical enough" whereas if your xG conceded is higher than the actual, you are "lucky".  There can be various reasons for the difference.  But xG definitely can help to focus on how your team is actually performing, when the scorelines themselves may be misleading.

 

Take the example someone mentioned.  Team A is on the attack for most of the game but is kept to a bunch of 0.1 likelihood chances which add up to an xG of 1 but it takes none of them.  Team B plays on the break and scores from a single chance say 0.5.  So Team B wins but xG would put Team A as the winner.  What do we learn from this ?  I'd say, first that normally the team creating constant low quality chances is more likely to score at least 1 of them, but if Team A can't create better chances than 0.1, it's not playing that much better than Team B.  Whereas often in football you get better chances on the break because the opposition is not set up defensively.  Most of the time, the team with 10x0.1 chances will beat the team with 1x0.5 chance, but sometimes not - that's football.

 

 

  • Like 1

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×
×
  • Create New...