Fever Dream (1 Viewer)
- Thread starter procdoc
- Start date
Danceswithhorses
Well-Known Member
Pretty remarkable day in our history, especially when compared to how bad it got under our last ownership.
oscillatewildly
Well-Known Member
The whole day did seem a little surreal.
Bristol sky blue
Well-Known Member
Wasn't at the game, but watching YouTube vids it seemed that lots of our fans were too shocked to celebrate the goals eventually... Surely we don't score goals as good as Torp's second, do we?!
skybluelee
Well-Known Member
Most of us were half laughing/half cheering by the time the 6th and 7th went in due to the sheer absurdity of it all.
According to the article yesterday in The Athletic, this was the 3rd largest discrepancy in xG to actual goals in over 18600 games across the major leagues where xG has been recorded. The odds of scoring from all of those 7 shots were calculated at 25000/1. Most of them had an individually very low xG.
In short, it was a freak outcome and doesn’t prove or disprove the concept of xG.
Robinshio
Well-Known Member
until we repeat it next week !! XG only works if you take into account the quality of the player given the specific chance
Torp will score that 1 in 3 - but not many others would
3.26 XG v Derby 5 goals
1.11 XG v QPR 7 goals
letsallsingtogether
Well-Known Member
My brother said to me when I dropped him off.
"Phone me when you get home to make sure this is not a dream. If it is what time you picking me up!"
Terry_dactyl
Well-Known Member
i don’t really understand it but it sounds quite subjective.
True, it doesn’t account for the actual player (how could it?). From memory, Torp’s second was put as 0.01 xG, effectively saying 1 in 100 shots from that place (with the defence set as it was) would go in.
xG stats are useful as a trend over time, not as specific instances.
viridisman
Active Member
If honest, id say 20,000 out of the 800 or so qpr fans that were still remaining cheered their goal.
It's gotta hurt when you score away and the home fans are making the loudest cheer...
Offhegoes
Well-Known Member
It wasn't a freak outcome. we were all over QPR.
The cheer from our fans for their goal was that loud that QPR fans that were walking away thought we'd scored another goal.
Ok, if we’re being pedantic, it was freak occurrence for 7 shots out of 8 to go in.
Have you ever seen that before, or anything like it?
ovduk78
Well-Known Member
Brazil 1 Germany 7?
JimmyHillsbeard
Well-Known Member
I’m afraid you imagined it all. You should have realised how preposterous it was when it involved a home game with a Saturday 3pm kickoff.
Try harder. Germany’s xG was 3.1 and they had 14 shots for 7 goals. Nowhere near comparable. Do you have a better example?
ovduk78
Well-Known Member
No, it was just a guess. I remember watching it and in the 1st half especially everytime Germany attacked they seemed to score. Brazil also had plenty of decent chances to score more than 1.
JimmyHillsbeard
Well-Known Member
Randomly as Cov had a cup match the next day, and I met up with a Dutch friend visiting from the US, I went to this 6 seasons ago. No idea about the xG (although I bet Luton’s was 0.0).
Brentford 7-0 Luton Town: Josh Dasilva hits hat-trick in Bees' biggest win in 25 years
Brentford record their biggest win in 25 years as Josh Dasilva scores a hat-trick in a 7-0 thrashing of struggling Luton Town.
www.bbc.co.uk
duffer
Well-Known Member
So I work a little bit with data and modelling (I'm more your plus size model, unfortunately), and I've got to say that I'm not convinced xG is particularly accurate.
Does xG account for things like player ability, surely a key predictive variable, or is it strictly based on pitch location and defensive set up.
Then you get to things like is it regularly re-validated and/or re-calibrated, or is it closer to a back of a fag packet metric than to a genuine model. It's not unfair to be sceptical.
Does xG account for things like player ability, surely a key predictive variable, or is it strictly based on pitch location and defensive set up.
Then you get to things like is it regularly re-validated and/or re-calibrated, or is it closer to a back of a fag packet metric than to a genuine model. It's not unfair to be sceptical.
RoysterE17
Well-Known Member
Keep watching replays and extended highlights of the game, everytime I've watch it i think what a performance by everybody!!
ovduk78
Well-Known Member
Apparently BTA's goal had an xG of 0.27, surely that is way too low. I wonder what the xG is for a penalty, surely above 0.5
Bertola
Well-Known Member
Apparently BTA's goal had an xG of 0.27, surely that is way too low. I wonder what the xG is for a penalty, surely above 0.5
There are lots of different models, but usually its around 0.7 - 0.8 for a penalty.
The atheletic actually did an article on us "breaking" xG at the weekend. Really good read.
Does Coventry 7 (xG 1.27) QPR 1 (0.61) prove expected goals is broken?
The Championship club's remarkable overperformance on Saturday raised questions about the xG model
www.nytimes.com
Sky_Blue_Dreamer
Well-Known Member
Similar to what I said in the QPR match thread and thought "the actual stats/data people are probably laughing at this because I've no idea how the actual metric works".
Nice to know someone that has some knowledge in the subject wonders the same thing.