Page 1 of 2 12 LastLast
Results 1 to 10 of 13
  1. #1
    Members
    Join Date
    Dec 2011
    Posts
    2,629

    Experimental 3-6-1: Scatter Graphics PL so far


    2 Not allowed!
    One drawback in being in the top league is that the brains behind Experimental 3-6-1 generally concern themselves more with the Football League. However, they have produced scatter graphs on various stats for the PL so far this season. It makes for interesting, yet still not clarifying reading on the new approach so far:

    https://experimental361.com/2019/09/...e-17-sep-2019/

    In terms of shot dominance, our possession based approach puts us in the top half. However, as we all have witnessed, our attacking effectiveness needs improvement. We're in the bottom 5 for shots taken per match and the bottom 8 for the number of shots it takes us to score a goal. Our defending seems a better story. We are in the top half for the number of shots we are facing per match. I would venture to suggest that the poorer looking stat for the number of shots it takes for a team to score against us is slightly skewed by the clinical Man City strikers' impact upon a very small sample data measured over just 5 matches.

    My reading would suggest that this supports what we all know anyway. We seem to be on the right track, but its very early days. We obviously need to be creating more chances. Hopefully that will come with time.

    The weird readings of these tables seem to be those of Watford and Norwich. Their stats could suggest that they should be in each other's league position. It will interesting to see if Watford start to get better results and whether Norwich can maintain things. The quality of the chances would obviously have an impact, but you could argue that, should they have each other's strikers, Watford would be top four and Norwich would be doomed already.

    • North Stand Chat

      advertising
      Join Date: Jul 2003
      Posts: Lots

        


    • #2
      Members
      Join Date
      Oct 2013
      Location
      Hanover/Elm Grove
      Posts
      9,590


      0 Not allowed!
      Thanks as ever. My first impression of this article was a pop-up (or whatever it's called), and I was expecting some request to consenting to using cookies but, instead, it indicated that they will be taking part in the climate strike on Friday and wanted to see if I'm in
    • #3
      Members
      Join Date
      Oct 2013
      Location
      Hanover/Elm Grove
      Posts
      9,590


      0 Not allowed!
      Turning now to the analysis, really interesting stuff, and it reinforces my initial impressions about Potter. Alongside the Norwich-Watford outliers, I'd add that I'm really surprised at how well Sheff Utd are doing and, if it continues, this analysis suggests they might well survive.
      On the basis of this, it's Norwich, Newcastle, Villa and Palace that ought to be most concerned.
    • #4
      Members
      Join Date
      Dec 2011
      Posts
      2,629


      0 Not allowed!
      Quote Originally Posted by Machiavelli View Post
      This quote is hidden because you are ignoring this member. Show Quote
      Turning now to the analysis, really interesting stuff, and it reinforces my initial impressions about Potter. Alongside the Norwich-Watford outliers, I'd add that I'm really surprised at how well Sheff Utd are doing and, if it continues, this analysis suggests they might well survive.
      On the basis of this, it's Norwich, Newcastle, Villa and Palace that ought to be most concerned.
      Sheff Utd seem to be our closest comparator so far. We are near each other in all these tables. We have both played only one of the top six, we both lost at home to Southampton in a match where we both had a player sent off. We also seem to be experimenting with their 'over-lapping centre halfs' (but it's all anyone in the media says about them and nobody seems to have noticed us doing it.) We are on the same points and next to each other in the table. I would suggest that the that we may have the greater potential to improve, and that they are still on a promotion bounce, but obviously that could be wishful thinking on my part. To be honest, I've really no idea whether a 'promotion bounce' is a statistically viable proposition, or just a pundit's cliche.
    • #5
      Unused sub REMF 4 Yoda's Avatar
      Join Date
      Mar 2004
      Location
      Worthing, West Sussex, United Kingdom
      Posts
      9,020


      1 Not allowed!
      The expected goals based table proves to show how unlucky we feel we've actually been
      Why DON'T YOU ALL JUST
      Forza la Viola, Forza Fiorentina!!!
      Worthing Raiders RFC!!!
    • #6
      Members
      Join Date
      Oct 2013
      Location
      Hanover/Elm Grove
      Posts
      9,590


      0 Not allowed!
      Quote Originally Posted by Stato View Post
      This quote is hidden because you are ignoring this member. Show Quote
      Sheff Utd seem to be our closest comparator so far. We are near each other in all these tables. We have both played only one of the top six, we both lost at home to Southampton in a match where we both had a player sent off. We also seem to be experimenting with their 'over-lapping centre halfs' (but it's all anyone in the media says about them and nobody seems to have noticed us doing it.) We are on the same points and next to each other in the table. I would suggest that the that we may have the greater potential to improve, and that they are still on a promotion bounce, but obviously that could be wishful thinking on my part. To be honest, I've really no idea whether a 'promotion bounce' is a statistically viable proposition, or just a pundit's cliche.
      Again, all interesting. I'd add that, alongside their promotion bounce, there is our adaptation to a drastically different way of playing, whereas they've had continuity on that front.
    • #7
      Members
      Join Date
      Apr 2011
      Location
      Beaminster, Dorset
      Posts
      2,642


      0 Not allowed!
      Quote Originally Posted by Yoda View Post
      This quote is hidden because you are ignoring this member. Show Quote
      The expected goals based table proves to show how unlucky we feel we've actually been
      No, it doesn't. It shows how many times we have got the ball in areas that statistically are shown to result in goals. End of. Fact is that we don't convert enough. Well there's a thing, sure GP will be sitting at home saying he never knew that.

      Secondly, the table above converts the hypothetical into points, which is close to meaningless after 5 games as you are looking at a hypothetical indicator for two teams, then extrapolating into points for that game. It is beyond meaningless for such a small sample.

      Look the table up for whole of last season and you will see there is a good correlation between Xg and final position (although Albion should have been relegated behind Cardiff) up to top 8 or so teams where the quality of strikers makes a huge difference to the conversion rates such that their actual points are considerably higher than Xg points. All Xg does is emphasise how important good strikers are. We have one pensioner, a guy from Championship, and a 19 year old with two minutes' PL experience, having sent out Kemy 2nd and Bruce Lee to pastures foreign. Not sure how that works.
    • #8
      Members
      Join Date
      Jan 2010
      Posts
      7,976


      1 Not allowed!
      Quote Originally Posted by warmleyseagull View Post
      This quote is hidden because you are ignoring this member. Show Quote
      No, it doesn't. It shows how many times we have got the ball in areas that statistically are shown to result in goals. End of. Fact is that we don't convert enough. Well there's a thing, sure GP will be sitting at home saying he never knew that.

      Secondly, the table above converts the hypothetical into points, which is close to meaningless after 5 games as you are looking at a hypothetical indicator for two teams, then extrapolating into points for that game. It is beyond meaningless for such a small sample.

      Look the table up for whole of last season and you will see there is a good correlation between Xg and final position (although Albion should have been relegated behind Cardiff) up to top 8 or so teams where the quality of strikers makes a huge difference to the conversion rates such that their actual points are considerably higher than Xg points. All Xg does is emphasise how important good strikers are. We have one pensioner, a guy from Championship, and a 19 year old with two minutes' PL experience, having sent out Kemy 2nd and Bruce Lee to pastures foreign. Not sure how that works.
      Wow. Grumpy or what.
    • #9
      Members
      Join Date
      Oct 2013
      Location
      Hanover/Elm Grove
      Posts
      9,590


      0 Not allowed!
      Quote Originally Posted by Plooks View Post
      This quote is hidden because you are ignoring this member. Show Quote
      Wow. Grumpy or what.
      I'm struggling to identify what the 'what' might entail
    • #10
      Members
      Join Date
      Aug 2008
      Location
      The land of chocolate
      Posts
      1,077


      0 Not allowed!
      I like this graph from https://twitter.com/S2Stats/status/1173925493629956097. Points to a comfortable season assuming our shortfall in goals vs expected goals is down to random variation rather than having a team who are bad at finishing.

      Going by this, Arsenal could be in for a shocking season if Aubameyang gets injured or their defence doesn't improve.

      Click image for larger version. 

Name:	xGxGA.PNG 
Views:	177 
Size:	124.5 KB 
ID:	115410
      Don't make me run. I am full of chocolate.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •