Wor Eddie's Fuck The Draw Bastards v Eriks Schmoking the Pancake School Meal Wankers - AKA The League Cup Final + Gemmill Loves Excel

[Toonpack 9385]

3 minutes ago, The Fish said:

Of the 220 PL games played so far, 137 games have gone the way of xG. That's 62%.

51 of the remaining 83 have <0.7 variance. (In other words 186 results have been at least close to the 'expected' result)

 

So there've only been 32 results from the Premier League season that have gone differently than the xG suggested, by a significant measure.

 

Of those 32, 7 wins went to the team with the significantly lower xG. 

 

tl;dr I don't think you get it.

 

 

 

Tell you what, I reckon 100% of 220 prem league games this so far this season went the way of actual goals scored (as seen by the crowd with their eyes).

 

What else is there to get, some stat that can indicate to me the result 62% of the time, well fuck me, that's astonishing.

 

It's worthless man !!

 

 

[Toonpack 9385]

2 minutes ago, Gemmill said:

 

I mean it's not intended for you to tot it up and celebrate high xG chances during the game.

 

Flailing tbh. 

 

So what is it's purpose then ?? What's does it bring to the understanding/enjoyment of the game.

[The Fish 10849]

7 minutes ago, Dr Gloom said:

 

xG suggests we created the higher value opportunities, no? 

In terms of xG, the difference between the two sides was nominal. Put it this way, in xG terms the expected outcome of a penalty is around 78%, or 0.78 xG. So a variance of <0.25 is nothing. 

[Gemmill 44804]

5 minutes ago, PaddockLad said:

No one can agree on how to interpret xG?  What a completely unexpected surprise

 

Yeah, in the same sense that normal people and flat earthers can't agree on the shape of the planet.

 

[Gemmill 44804]

1 minute ago, Toonpack said:

 

So what is it's purpose then ?? What's does it bring to the understanding/enjoyment of the game.

 

For you? Nothing. 

[Meenzer 15518]

Well some people certainly have an xGIR (xG Interpretation Rightness) of 0.0, don't they?  

[Toonpack 9385]

3 minutes ago, Gemmill said:

 

For you? Nothing. 

 

No, for you ??

[Gemmill 44804]

UNSUBSCRIBE

[The Fish 10849]

3 minutes ago, Toonpack said:

 

Tell you what, I reckon 100% of 220 prem league games this so far this season went the way of actual goals scored (as seen by the crowd with their eyes).

 

What else is there to get, some stat that can indicate to me the result 62% of the time, well fuck me, that's astonishing.

 

It's worthless man !!

 

 

 

It's not about telling the results man, it's providing context. So, our home game against Leeds, finished 0-0, right? So what xG tells us is that we actually created enough good chances to score at least 1 goal, possibly even 2. With 2 additional numbers you immediately learn a little more about the game. 

 

A casual fan might see 2-0 and assume Man Utd were much the better team. But xG and our eyes tell a different story, don't they?

 

I don't understand why you're so anti something that shouldn't impact your enjoyment of football one iota? You don't care about PPDA, or ODC, so why do you care about xG?

 

[Meenzer 15518]

5 minutes ago, The Fish said:

You don't care about PPDA, or ODC, so why do you care about xG?

 

 

I'll have you know I don't care about any of the Wu-Tang Clan.

 

[PaddockLad 17240]

15 minutes ago, Gemmill said:

 

Yeah, in the same sense that normal people and flat earthers can't agree on the shape of the planet.

 

What does Eddie Howe do with his xG stats to improve the team?  

[Toonpack 9385]

6 minutes ago, The Fish said:

 

It's not about telling the results man, it's providing context. So, our home game against Leeds, finished 0-0, right? So what xG tells us is that we actually created enough good chances to score at least 1 goal, possibly even 2. With 2 additional numbers you immediately learn a little more about the game. 

 

why do you care about xG?

 

 Point #1 - Re Leeds, yes I saw all that watching it with my eyes without the need for any numbers.

 

Point #2 - I don't, but you and Gemmill don't half jump about.

 

 

[Dr Gloom 21912]

10 minutes ago, The Fish said:

 

It's not about telling the results man, it's providing context. So, our home game against Leeds, finished 0-0, right? So what xG tells us is that we actually created enough good chances to score at least 1 goal, possibly even 2. With 2 additional numbers you immediately learn a little more about the game. 

 

A casual fan might see 2-0 and assume Man Utd were much the better team. But xG and our eyes tell a different story, don't they?

 

I don't understand why you're so anti something that shouldn't impact your enjoyment of football one iota? You don't care about PPDA, or ODC, so why do you care about xG?

 

 

i'm surprised we have a higher xG than man u for the game on sunday tbh. are you not?

 

xG suggests a 1-1 draw might have been about right given the quality of the chances both sides created in 90 minutes. we controlled possession and had them on the back foot for large parts of the game but to the naked eye, it looked to me like they created the better chances and it wasn't just a case of them being clinical. 

 

i'm not against xG in principle but who crunches the numbers? who does the modelling? i've noticed that they're often revised up or down after the game. whoever was in charge on sunday seemed about as good at their job as the muppets operating VAR...

[ewerk 30545]

6 minutes ago, Dr Gloom said:

 

i'm surprised we have a higher xG than man u for the game on sunday tbh. are you not?

 

Not really and I watched the match with MY EYES. It was a game of very few clear cut chances and the xG bears that out. Casemiro's goal was the biggest chance of the match, other than that neither side really created anything of note.

[The Fish 10849]

14 minutes ago, PaddockLad said:

What does Eddie Howe do with his xG stats to improve the team?  

 

Well, each chance is given a xG score, so he'd be able to see what patterns of play produce the best value chances. Something that worked really well in the early stages of the season was our work down the right hand side. The interplay between trippier, Almiron and Bruno was creating multiple decent chances each game. The issue we're facing now is that Trippier hasn't been as superb as he was, Almiron's purple patch is fading to a hardworking but less brilliant violet and Bruno's been suspended. 

 

He'll be aware that we're not taking a great deal of shots despite creating a good number of opportunities so he can instruct the players to take more shots from lower xG opportunities. 

 

He'll be aware of the players contributing the most to passages of play that end up with a big chance and will work to get them more involved in the game, or tweak another's' role to increase their threat.

 

y'know, stuff like that

[PaddockLad 17240]

3 minutes ago, The Fish said:

 

Well, each chance is given a xG score, so he'd be able to see what patterns of play produce the best value chances. Something that worked really well in the early stages of the season was our work down the right hand side. The interplay between trippier, Almiron and Bruno was creating multiple decent chances each game. The issue we're facing now is that Trippier hasn't been as superb as he was, Almiron's purple patch is fading to a hardworking but less brilliant violet and Bruno's been suspended. 

 

He'll be aware that we're not taking a great deal of shots despite creating a good number of opportunities so he can instruct the players to take more shots from lower xG opportunities. 

 

He'll be aware of the players contributing the most to passages of play that end up with a big chance and will work to get them more involved in the game, or tweak another's' role to increase their threat.

 

y'know, stuff like that

I think he’ll be able to do that by just watching the games mate. Someone mentioned lack of shots in this thread earlier or yesterday and I agree, neither of us resorted to xG, we didn’t need to. This is the point, xG is 80% observable. Yes anomalies to prove underlying points will occur but not enough to make it that valuable. You lot can’t agree today on it and KSG posted an article doubting its effectiveness too. Shall we just agree it all a load of fuckin old bollocks? 

[Dr Gloom 21912]

7 minutes ago, ewerk said:

 

Not really and I watched the match with MY EYES. It was a game of very few clear cut chances and the xG bears that out. Casemiro's goal was the biggest chance of the match, other than that neither side really created anything of note.

I’ll have to rewatch it 

 

[The Fish 10849]

9 minutes ago, Dr Gloom said:

 

i'm surprised we have a higher xG than man u for the game on sunday tbh. are you not?

 

xG suggests a 1-1 draw might have been about right given the quality of the chances both sides created in 90 minutes. we controlled possession and had them on the back foot for large parts of the game but to the naked eye, it looked to me like they created the better chances and it wasn't just a case of them being clinical. 

 

i'm not against xG in principle but who crunches the numbers? who does the modelling? i've noticed that they're often revised up or down after the game. whoever was in charge on sunday seemed about as good at their job as the muppets operating VAR...

 

I'd agree a low scoring draw would be about right. The two chances they scored from weren't their best. Casemiro's chance goes begging 9 times out of 10, as we saw when a very similar chance came for us in the second half. Roughly the same position for the freekick, put into the same area, but this time it didn't come off. 

 

I'm reminded of a goal Wilson scored for us earlier on this season. Can't remember who it was against, but the ball came in from the left and he ran onto it at the near post and prodded it at an obtuse angle to the far corner of the net. That chance was something like 0.1 xG, but the finish was so good, it had a post-shot-xG of something like 0.9. When I talked about it later on I was surprised how many people described it as a good chance because it really wasn't. The angle was bad, he had a defender on him, the 'keeper was in a decent position and I'm not even sure it was with his favoured foot. But because it resulted in a goal, people said it was a good chance. 

 

There are competing models of xG, which is good, because it means they're trying to improve the reliability of the data. OPTA are pretty good though. 

 

[The Fish 10849]

8 minutes ago, PaddockLad said:

I think he’ll be able to do that by just watching the games mate. Someone mentioned lack of shots in this thread earlier or yesterday and I agree, neither of us resorted to xG, we didn’t need to. This is the point, xG is 80% observable. Yes anomalies to prove underlying points will occur but not enough to make it that valuable. You lot can’t agree today on it and KSG posted an article doubting its effectiveness too. Shall we just agree it all a load of fuckin old bollocks? 

 

Because our eyes lie all the time. Confirmation bias, outcome bias, it's always there for everyone of us.

 

xG isn't replacing anything, it's just providing context, adding more detail. 

 

The only thing that is bollocks is people reading that 'article' by KCG and thinking it's damning. It's like a creationist dismissing evolution because it's 'only a theory'. That article says xG isn't perfect, but says that as more data points are added, as greater granularity is added, it will become even more accurate. And even in it's current state, I'd trust an xG stat before I listened to some gobshite on here.  

Edited February 28, 2023 by The Fish

[PaddockLad 17240]

2 minutes ago, The Fish said:

 

Because our eyes lie all the time. Confirmation bias, outcome bias, it's always there for everyone of us.

 

xG isn't replacing anything, it's just providing context, adding more detail. 

 

The only thing that is bollocks is people reading that 'article' by KCG and thinking it's damning. It's like a creationist dismissing evolution because it's 'only a theory'. That article says xG isn't perfect, but says that as more data points are added, as greater granularity is added, it will become even more accurate. And even in it's current state, I'd trust an xG stat before I listened to some gobshite on here.  

People are saying stuff that is borne out in the stats without the need to resort to them though Dave

 

Its not bollocks, but it is largely irrelevant….

[The Fish 10849]

36 minutes ago, Toonpack said:

 Point #1 - Re Leeds, yes I saw all that watching it with my eyes without the need for any numbers.

 

Point #2 - I don't, but you and Gemmill don't half jump about.

 

 

 

What about Southampton 1 - 3 Brighton? or West Ham 2-0 Bournemouth?

[The Fish 10849]

1 minute ago, PaddockLad said:

People are saying stuff that is borne out in the stats without the need to resort to them though Dave

 

Its not bollocks, but it is largely irrelevant….

 

No, people are saying unmitigated horsepiss, that is refuted by the stats and personal testimony. 

It used to be the case that Jimmy said X and Tyrese said Y and the two of them could argue back and forth forever. Now, Tyrese shows a stat that backs up his claim so Jimmy can stfu orrrrr Jimmy can dismiss the validity of the stat and they can have a circular argument about that instead.

 

It's neither bollocks, nor irrelevant. It's an integral part of the professional game now, both in coaching, scouting and punditry and yes, even arguments on messageboards.

[Toonpack 9385]

3 minutes ago, The Fish said:

 

What about Southampton 1 - 3 Brighton? or West Ham 2-0 Bournemouth?

 

Hmmmm does it mean Brighton and West Ham won ???.

[The Fish 10849]

2 minutes ago, Toonpack said:

 

Hmmmm does it mean Brighton and West Ham won ???.

Yeah, but what did your eyes tell you about the games, were they fair results? Were there many chances in each? Were Bournemouth lucky to get the 0? Howay, what do your eyes tell us about those games?

 

Ya Gonk.

[trophyshy 7083]

This is worse than losing the final.