Cooper - OUT

[ClaphamFox]

6 minutes ago, Stadt said:

I'm not a fan of expected points, preferring to use xG, xA and xGD. With expected points you have to make a very arbitrary calculation that say .3 between two teams is a draw and more than that is a win.

 

There's a reason the analytics departments of the world's best and smartest clubs lean on it so heavily. I'd really recommend How to win the premier league, Ian Graham, Liverpool's director of research's book

Thanks - I'm genuinely curious to know why he thinks Liverpool managed to exceed their xPts by 25 when they won the league in 2020.

[Gamble92]

2 minutes ago, ClaphamFox said:

Thanks - I'm genuinely curious to know why he thinks Liverpool managed to exceed their xPts by 25 when they won the league in 2020.

Because it makes him look better?

[Stadt]

3 minutes ago, ClaphamFox said:

Thanks - I'm genuinely curious to know why he thinks Liverpool managed to exceed their xPts by 25 when they won the league in 2020.

Because they have good players and coaches their analysts helped identify by looking at their stats, predicated mostly on xG.

[Babylon]

59 minutes ago, Finnegan said:

 

Months before, in fact the season before, we got relegated there were those of us arguing with the likes of Dahnsouff and probably yourself that we were starting to create troublingly few chances and that our xG plummeting and PPG sliding was going to be a problem longer term. 

 

Most of the same people downplaying our concerns then are in this thread now saying everything is fine because we're outside the relegation zone and we've picked up points in spite of the underlying stats. 

 

Hmmmmm consdiering you'd have to go back to mid 2001 to find me saying anything remotely nice about Rodgers or defending him I wouldn't be so sure. 

 

Neither have I said we'll be ok, merely pointing out Claphams post the other day which clearly showed few over the course of a season perform to their "expected level", could be more, could be worse. 

[FOXYTALK]

 

[CosbehFox]

29 minutes ago, Babylon said:

Hmmmmm consdiering you'd have to go back to mid 2001 to find me saying anything remotely nice about Rodgers or defending him I wouldn't be so sure. 

 

Neither have I said we'll be ok, merely pointing out Claphams post the other day which clearly showed few over the course of a season perform to their "expected level", could be more, could be worse. 

I did explain why this occurs on the expected points model. 

 

The best use of the x model is always how does this compare to others. 

[CosbehFox]

43 minutes ago, ClaphamFox said:

Thanks - I'm genuinely curious to know why he thinks Liverpool managed to exceed their xPts by 25 when they won the league in 2020.

I told you a couple of days ago, the xPts does not directly translate to points. You can get 2.2 expected points for a draw for example 
 

The bigger stat there is they got the most xPts and the most actual points. Therefore the model is correct 

Edited 9 October 2024 by CosbehFox

[Dahnsouff]

1 hour ago, Finnegan said:

 

Months before, in fact the season before, we got relegated there were those of us arguing with the likes of Dahnsouff and probably yourself that we were starting to create troublingly few chances and that our xG plummeting and PPG sliding was going to be a problem longer term. 

 

Most of the same people downplaying our concerns then are in this thread now saying everything is fine because we're outside the relegation zone and we've picked up points in spite of the underlying stats. 

 

Could at least name check me during your slandering Mr @Finnegan  as I do not dispute it,  and admit that I was wrong and was far too late to lambast what Rodgers had become and taken us with him.

 

Maybe that explains why I made no such statements this time. I may contest the nature of some points, stats are useful sometimes but  used to prematurely (in my opinion) using them to beat someone with, less so.

However, I do agree that the low rate of chance creation is a massive concern, and a point I agreed with the @The Doctor on, if it continues beyond these winnable games, we will likely be undone.

[CosbehFox]

The common misunderstanding with the x models is that the figures in isolation are a judge of performance. Their best use is in comparison against others performance, that’s why the points @Stadt made are salient. 
 

The xG model for example is subjective. You can probably find three different xG score lines for Saturday’s game for example. One shot might be marked 0.65, another source might mark it 0.48. So in isolation such stats as actual goals scored and xG are a subjective matter. You are rating performance against a subjective model. 
 

However if we use the same xG model for example and you compare other teams performance on that same scale. You have statistics which are more useful as a judge. 
 

When a professional club’s analyst identifies a player. Let’s say Jordan Ayew. They wouldn’t recommend them off an xG of 4.67 from a possible 6.15. They would recommend them in how Ayew’s performance compares to other players. You can see this with the fbref comparisons and percentile. The latter is more key driver in a signing but it’s all sourced by the comparison 
 

When we compare LCFC against other clubs on the x models, the performance is stinky. 

Edited 9 October 2024 by CosbehFox

[Dahnsouff]

Can see the benefit, but does this not mean viewing things "through a keyhole"? Suppose that is just an acceptance when utilising the statistics in question.

Edited 9 October 2024 by Dahnsouff

[ClaphamFox]

1 hour ago, CosbehFox said:

I told you a couple of days ago, the xPts does not directly translate to points. You can get 2.2 expected points for a draw for example 
 

The bigger stat there is they got the most xPts and the most actual points. Therefore the model is correct 

Not according to understat. According to their data, Liverpool's xPts in 2019/20 was 74.28 and Man City's was 86.76. So if we believe xPts, Man City should have won the title comfortably.

 

When we won the PL, Arsenal, Spurs and Man City all had higher xPts than us. When we got relegated in 2022/23, our xPts was better than seven other teams. 

 

 

 

 

Edited 9 October 2024 by ClaphamFox

[Sunbury Fox]

3 hours ago, ClaphamFox said:

Not according to understat. According to their data, Liverpool's xPts in 2019/20 was 74.28 and Man City's was 86.76. So if we believe xPts, Man City should have won the title comfortably.

 

When we won the PL, Arsenal, Spurs and Man City all had higher xPts than us. When we got relegated in 2022/23, our xPts was better than seven other teams. 

 

 

 

 

Aside from any stats demonstrating it or not, we can see with our own eyes that we're being outplayed for large periods of pretty much every game and conceding far more good chances than we're creating; I assume you're not disputing that? This being the case, if the same happens over a full season it is reasonable to expect us to lose a lot more games than we win and therefore the question is, will we be able to pinch results against the odds in enough games to secure the points we need to stay up? Doing that is probably going to rely on having a great keeper who saves a lot more big chances than the average keeper (tick) and a top striker who finishes a very high proportion of the few chances created (far less of a tick now that Vardy is well past his best).

Edited 9 October 2024 by Sunbury Fox

[orangecity23]

All this stat talk is somewhat missing the point - whether they are fully accurate or not, they are only a part of a bigger picture that says we are being outplayed for large portions of matches. You don't need a comprehensive stat analysis to tell you that we offered nothing in attack in the Spurs 1st half, the Fulham 2nd Half, the Villa 1st half, the Everton 1st half, the Arsenal 1st half, most of the Walsall game and the Bournemouth 2nd half. The stats are just another way of illustrating this - we give up a boat load of chances, and create very little. Cooper's tactics offer barely any margin for error, and where we are now in the table shouldn't be taken for granted or used as a whitewash to claim everything is fine. After the Arsenal game, a heroic performance from Hermansen to preserve our goal difference was the only thing separating us from the bottom 3, and likewise, on Saturday, it was our post and bar in the second half which prevented us from being down there still. That's how thin the margins are when you set up like this, because the "pragmatic" bit of camping in your own half only invites trouble as the opposition can commit every one forward to liberally pepper our goal at will. We actually look a lot better and defensively sound, and concede less goals when we are playing in "handbrake off" mode.

[CosbehFox]

4 hours ago, ClaphamFox said:

Not according to understat. According to their data, Liverpool's xPts in 2019/20 was 74.28 and Man City's was 86.76. So if we believe xPts, Man City should have won the title comfortably.

 

When we won the PL, Arsenal, Spurs and Man City all had higher xPts than us. When we got relegated in 2022/23, our xPts was better than seven other teams. 

 

 

 

 

As I’ve said before in this thread, it’s used largely by clubs to track a trend
 

Man City since that season go onto four league titles in a row. 
 

It’s like the seasons when Brighton finished 16th but 5th in xPts - that’s predicting an upturn which occurred over the next two seasons.  
 

Similar applies to Burnley when they finished 7th and all the xmodels were saying it wouldn’t last - mean performance returns the season after 

Edited 9 October 2024 by CosbehFox

[The Doctor]

7 hours ago, Babylon said:

But as he pointed out the other day and as he's alluding to, few teams perform to the "underlying numbers"

only if you apply a particular way of viewing the underlying numbers and also misapply it slightly. let's take last seasons xPts model and look not at the exact number of points but the average positional difference in actual table vs expected table: the average difference is 2.3 positions, driven massively by West Ham and Man Utd drastically over performing their expected position by 5 places each. unsurprising then that Man Utd have started pretty poorly.

 

the season before, 2.1 places out, driven by us underperforming by 6 positions and Fulham over performing by 6. Fulham then slipped back by 4 places the following season.

 

this trend tends to hold, the xTable vs the actual table on average has around a 2 position difference. So, while the points tally may not be wholly in agreement, the relative points tallies generally are more so - a team which on the underlying data is 19th are unlikely to be much beyond 16th/17th. You can say that well, 17th is all you need, but it's a pretty massive gamble that you'll get that level of outlying from the underlying data, and that it will go in your favour, and then what? you might outperform it for a season, but it's a reasonable predictor of ongoing performance unless you make significant changes 

 

[orangecity23]

Any stat has a margin of error on it in either direction. If it was a midtable stat and we were underperforming it a bit, or overperforming it it wouldn't be a big issue. The issue is that all of the base stats are abysmal. It means we HAVE to overperform it to deliver the heady heights of 17th, because all it will take is a little injury to Mads and a few games of Danny Ward, or a Vardy injury meaning we are relying on 3 goals a season Ayew up front and we will be right in the shit.

Edited 9 October 2024 by orangecity23

[TheLittleBigMan]

Yeah, here's a chart.

 

[kingpower]

Doesn’t feel like a particularly ‘legit’ source…

Edited 9 October 2024 by kingpower

[Tuna]

Messed it up

Edited 9 October 2024 by Tuna

[Spiritwalker]

One positive I’d say about Cooper is that he seems to be really lucky.

The season that he kept Notts Forest up they seemed to get battered

every week but sneak a result somehow. I wonder what their Xg was.

[pmcla26]

8 minutes ago, kingpower said:

Doesn’t feel like a particularly ‘legit’ source…

I think that’s just someone gauging fans feelings and looking at performances. Classic case of 2+2 etc. 

 

We all know he’s under pressure to succeed but that Bournemouth win has no doubt given him time. Obvious to us all as fans though that we’ve regressed from last season. 

[ClaphamFox]

19 minutes ago, kingpower said:

Doesn’t feel like a particularly ‘legit’ source…

It’s also stating the obvious. “We are told his position could come under threat if City suffer defeat at Southampton after the international break.”

 

Anybody could have told them that.

Edited 9 October 2024 by ClaphamFox

[winteriscoming]

I want Cooper out and this is no defence of him but I really have no interest in what the xg is for us in games.  It’s a very spursy stat. I’ll watch our games and judge it on how I see it not what our xg is or the opposition we are playing. We we’re very lucky to beat Bournemouth. Second half they could of been 4 or 5 - 1 up. I don’t need xg to tell me if we’re playing well or shit. 

[pmcla26]

I don’t necessarily want him out as I don’t have any dislike for the bloke, the feeling is nothing. 
 

I just want him to see the obvious. 
 

Annoyed he didn’t see the obvious in the summer when he directed us to sign Ayew and BDCR, too. 
 

Glover needs to take some responsibility too for deviating from the plan, although he’s not the one sanctioning the deals so there’s multiple people behind the scenes at fault. 

[Mark]

Graeme Bailey / GRV media are primarily an ad network, they provided ours here for 3/4 years. I doubt there is anything in that article unfortunately