Stat Attack

Below is the running record for xG across the Premier League season:
1685381736606.png
These figures shifted up after the first 8 matches (ending with the home draw with Brentford.) The two points up around 2.5 are statistical outliers, with the second one (home win against Fulham) marking the beginning of another shift up. The final four losses marked a return to anemic offensive punch. We finished 18th in the EPL for xG (only Saints and Wolves were worse.) However, the average xG during our good run starting with Fulham at home is about 1.5 per match, which would be more like a top 8 pace.

Here is the running record for xG against (xGA):
1685382141477.png
This series includes no statistical signals like those found in the xG series. That concerns me as it suggests that we indeed never did sort out our set piece defending. We finished 17th in the EPL for xGA (Forest, Everton, and Leeds were worse.)
 
Was reading up on some other stats that are the next layer along and one that really interested me was xT (I think expected threat).

It divides the pitch into squares and assigns a "threat" value between zero and one. The value is based on the chance of scoring within the next 5 actions from that position.

One interesting use of this is to measure particularly midfield players and their ability (or not) to progress the ball up the pitch to higher-value areas by taking the value they receive the ball in and subtracting it from the value of the position that the ball ends up in.

We have loads of discussions about midfielders Cook, Lerma and Rothwell in particular. I'd love to be able to see the figures for those guys as I think it would make for a really interesting debate.

Historical data on Surman would be good too.

Other than a few news articles I can't find the data anywhere.
 
Kudos said:
Was reading up on some other stats that are the next layer along and one that really interested me was xT (I think expected threat).

It divides the pitch into squares and assigns a "threat" value between zero and one. The value is based on the chance of scoring within the next 5 actions from that position.

One interesting use of this is to measure particularly midfield players and their ability (or not) to progress the ball up the pitch to higher-value areas by taking the value they receive the ball in and subtracting it from the value of the position that the ball ends up in.

We have loads of discussions about midfielders Cook, Lerma and Rothwell in particular. I'd love to be able to see the figures for those guys as I think it would make for a really interesting debate.

Historical data on Surman would be good too.

Other than a few news articles I can't find the data anywhere.
Click to expand...

I've not seen the figures either. I know there are figures about attacking threat, which does make sense. Repeated balls flashed across a box just evading an attacker get an xG of 0, but we all know it could have been very close to a goal. For us oldies, think Gazza vs Germany in 1996.
 
You must log in or register to reply here.

;