The relationship between xA and xG

One of the biggest problems in science is unreported negative or null findings. It leads to time wasted on replicating failed trials.

So with that in mind I'm going to blog about expected assists as a percentage of expected goals and why it isn't really enough to tell us anything.

My hypothesis was that we'd see the patient build up teams having relatively high xA as a percentage of xG and the longer ball teams with a low percentage.

I took the stats from Understat, stuck them into a spreadsheet and got the following results:

It looks OK.

Arsenal and Chelsea are high possession teams.

Cardiff and Crystal Palace more direct teams.

They are about where I thought they would be.

But then Manchester City are sandwiched between Huddersfield and Newcastle, and Liverpool and Manchester United are creating a lot more unassisted xG than my hypothesis would suggest.

My theory now is that high pressing teams like Liverpool will generate unassisted xG from forcing mistakes or turnovers. Manchester City probably do as well (like the Barkley assist at the weekend) which probably moves them leftwards in the table despite also having a lot of assisted xG.

So there we go, nice idea, but probably doesn't really tell us anything that useful.


Comments

  1. Andrew BeasleyFebruary 18, 2019 at 4:27 AM

    Did you strip out penalties? There's some very high value unassisted xG right there...

    ReplyDelete
    Replies
    1. @sbunching February 18, 2019 at 6:52 AM

      That chart is all goals, when I repeated for non-penalty the major mover is Manchester United (who seem to have had a lot of penalties). Doesn't xA accumulate on some of the events that result in penalties though?

      Delete
    2. Andrew BeasleyFebruary 19, 2019 at 2:55 AM

      I don't think Understat's xA includes penalty stuff, no. If you select a penalty on a match page, the assist section is blank.

      Delete

Post a Comment

Popular posts from this blog

Wyscout review and poking around the French third tier

Image
I signed up for Wyscout on Friday and was given full access for 48 hours. Inbetween the busy bits of family life over a weekend in the run up to Christmas I thought I'd do a quick review of it and reveal some of the players I liked the look of. For those who have never used Wyscout it actually looks a lot like Football Manager (or at least Football Manager about 10 years ago) The home page allows you to select a league to scout. The package I was trialing allowed me full access to 5 leagues (I chose the big 5) and partial access to all the others. To be honest I really couldn't see that much difference between full/partial access for player scouting purposes. I decided to use the trial time to see if I could use their stats and video clips to find some promising players in obscure leagues. So I looked in the French 3rd division, my thoughts being that France is producing so much talent that their third level must have some potential top quality players in it - with a t...
Read more

Scouting report Dan Ndoye - Lausanne Sport

Image
I've decided to try out a new approach to scouting. Instead of looking at how well the player is playing which almost always shows up positively in the data I am going to look at the attributes the player has. I'm going to look at a combination of data and video and record the results here.  The categories I've chosen for assessing the player aren't ones I've given a huge amount of thought to, it is an initial list of the things I think are important to predict the long term career progression of young players. Player - Dan Ndoye Club - Lausanne Sport Age - 18 Position - plays across the front 3, most often as an inverted winger. Technical Foot Good ball striking ability with either foot. Comfortable using either to shoot with and generates good power. Probably gets a slightly better shape on crosses from the left but has played wide on either side of the attack and tends to cut in and shoot more than cross. Touch Close control with either foot ...
Read more

Data Analytics conference - Daniel Krueger report

Image
Seattle Sounders Inaugural Analytics Conference (6/22 – 6/23 2019) Summary by: @DanielKrueger96 Organizers: @analysefooty @SSFCSpSci #SSFCAnalytics Day 1 Opening Remarks Garth Lagerway, President of Soccer – Seattle Sounders FC Building an analytics department from scratch Starting with current roster: sports science emphasis (RPE, sleep, athlete surveys, performance testing Player recruitment Inherently less data on players outside of your own club because other clubs don’t release sports sci data Building automation Finding objectivity to allow for data driven decisions Introduced Ravi Ramineni – Director of Analytics, Seattle Sounders FC General takeaways: There seems to be a common theme with sports front offices. Team presidents and GMs rarely have a data background. The role of an in-house analyst is fairly new, so this makes sense. Though front office management may not have an understanding of the technical side of the analyst’s day to do task, it is important...
Read more