CL Round: Newcastle - Barca 1-2

[Maradona37]

Long story short:
xG of 1 does not equal one good chance.
If you think that an xG of 0.3-0.4 translates to an actual sitter, you put that into different perspective

The way I thought it worked was a sitter would be like 0.8, is that not true?

I do get your point, but it doesn't seem to account for games where both teams seem to have an equal amount of good or slim chances, one team has an xG of almost three higher.

Anyway not too bothered - it's a decent tool but obviously has some flaws too and is in its infancy and yet to be (almost) perfected. It must be used in addition to other stats and the eye test. I hope you think that's a fair statement.

 

[Birdy]

The way I thought it worked was a sitter would be like 0.8, is that not true?

I do get your point, but it doesn't seem to account for games where both teams seem to have an equal amount of good or slim chances, one team has an xG of almost three higher.

Anyway not too bothered - it's a decent tool but obviously has some flaws too and is in its infancy and yet to be (almost) perfected. It must be used in addition to other stats and the eye test. I hope you think that's a fair statement.

No unfortunately. xG 0.8 is practically an empty net.
The fact that many people had missed (and will continue to miss) such empty nets makes it impossible to have it at 1 or close because of how math works in statistical probabilities

Then it is also the model. Trust Opta and Statsbomb

 

[Andresito]

Penalty is almost 0.8 no?

 

[Maradona37]

No unfortunately. xG 0.8 is practically an empty net.
The fact that many people had missed (and will continue to miss) such empty nets makes it impossible to have it at 1 or close because of how math works in statistical probabilities

Then it is also the model. Trust Opta and Statsbomb

Fair enough. I guess that makes sense - if plenty players miss sitters, the model must account for that. Makes you wonder how any teams build xGs of over 3 and 4 at all.

Maybe you'll say it is my anti Man United bias, but United seem to always do well in this stat, even in games I don't think they create particularly good chances. Equally, I see teams create good chances against them (like the night Spurs scored four, allowing one was from a corner) and they only have an xG of 0.78 lol.

I think it's a good tool and has its uses, but has some flaws as Jam points out and the numbers arrived at need to be questioned more at times rather than believing them on face value.

Just my take.