The on topic thread 2.0

Remove this Banner Ad

Status
Not open for further replies.
As drd23 pointed out, it might provide an indication for where Chelsea’s problems. This data suggests they can’t finish a sandwich and/or their goalkeeper has hands made of lettuce.

Finished ok against us. :(
 
Us in 11th sounds about right. Whilst the club's have access to better metrics, xG can really show a teams over or under performance.

At the halfway point of last season when we were going with city and Liverpool the xG table had us 11 points behind. The second half of the season it evened out and we won less points than the xG would have predicted. Any non one eyed watcher of Spurs games in the first half of the season would have seen a team getting points it's gameplay didn't deserve
 
A while back I read a really interesting article about how clubs were increasingly using these xG/xA metrics, from what I understand they’re becoming more sophisticated, certainly a lot more sophisticated than the models used by most publicly accessible football stats pages, and are highly valued and protected IP, but fundamentally they’re using the same principles. I’ll see if I can find it.

Essentially Liverpool (and I’m sure a number of other clubs) keep their own database which aims to document every shot by every player across various leagues in Europe and around the world, and using this real world data, estimate unique probabilities that players will score or create a scoring opportunity from various locations on the pitch under varying degrees of difficulty including shot angle and proximity of a defender.

For a particular player a shot in the 6 yard box under no pressure might be worth 0.95xG, a shot from 20 yards out on an angle maybe only 0.05xG. A different player or situation might be rated as 0.8 and 0.01xG. There’s a similar algorithm for assists (xA).

Sometimes teams and players will under or over perform relative to xG but over time they’ll regress to the mean.

Accumulate enough of this data over time for enough players and teams, and with the right people analysing it, you can build a reasonably accurate model to guide offensive and defensive strategy, analyse performances, identify under or overvalued players, etc.

People might not rate it as a metric but it’s definitely something the best clubs are valuing and using more and more. Similarly I suspect it’ll be something the footballing public will understand and appreciate more over time.
That's a really long way of saying that you downloaded City's scouting database.
 

Log in to remove this ad.

Not a surprise based on what I saw of those games. Finishing and goalkeeping obviously comes into it.
Not a lot of goalkeeping in the first game. The Shrewsbury game yes. I mean it's why Newcastle have 6 wins instead of 1. They are a bog awful team kept alive by a goalie and scoring scrappy goals when they need it. Probably have the lowest shots on goal.
 


👍👍👍

You just get the warm feeling the fairytale Leeds return to the top flight is gonna finally happen, with so many of our promotion rivals shitting the bed lately.

This club is not very good for the ticker though, but always immensely satisfy to beat the Sewer rats or Millwank 😆😆😆 #MOT
 

(Log in to remove this ad.)

Scum hanging with The Scum.

Sometimes a lack of patience is all a club needs. :)

images
 
Status
Not open for further replies.

Remove this Banner Ad

Back
Top