OMT ***Tottenham Hotspur v Manchester United***

[scaramanga]

But XG is just the distance from goal. Hence it is useful over 'many' games, but for one game can be misleading e.g. an open goal from 6 yards versus a rebound at your throat with 8 men on the line from 6 yards.

It's more than that. It takes into account where the ball came from, whether it was a headed chance or to feet, where defenders are, etc.

 

[Bullet]

It's more than that. It takes into account where the ball came from, whether it was a headed chance or to feet, where defenders are, etc.

Where defenders are? Are you sure?
I know they try to allow for foot/head and cross/pass but that is very subjective e.g. Alli's chance could have been a flying header, a cross from out wide from Trippier is also a pass etc

 

[scaramanga]

Where defenders are? Are you sure?
I know they try to allow for foot/head and cross/pass but that is very subjective e.g. Alli's chance could have been a flying header, a cross from out wide from Trippier is also a pass etc

https://www.google.co.uk/amp/www.te...-show-man-city-should-win-premier-league/amp/

Not everyone uses the same method of eGV, so it depends on where that came from. xG used to be very simplistic but most use a more nuanced model now AFAIK

 

[braineclipse]

The discussion is about who had the better chances in this one game though...

The sample size is what it is by virtue of that - compared to a vast sample size our XG in this game was better than there's - which pretty much sums up the 'who had the better chances in this game' debate

we just beat Madrid though so who cares...

And again my point all along was that just looking at the chances with the naked eye is more informative because of the small sample size.

Over this sample xG doesn't accurately present who had the better chances.