Scott Carson

[therealhantsram]

Opta’s review of 2019 ranks Scott Carson 24th out of 24 Championship keepers.  ?

I knew his performance wasn’t up to previous seasons but didn’t realise just how bad he was last term.

https://www.optasportspro.com/static/10-create-season-review.html

 

[Day]

This new x (expected) stat that’s being thrown around is the worst stat in football. 

I expect better from Opta. Bin it now.

It’s on par with the level of nonsense that is Sky Sports News using video game stats to compare players.

[irobinson]

He is a quality GK underneath!!

i was watching the goal keeping clip last night - and couldn’t help noticing how fitter / trim he looked 

 

lets hope he gets his sharpness back

 

Wouldn’t surprise me if he started as number 1 under Cocu - experience counts a lot in my book for a GK

[DontTrustTheDevilInside]

Even the headline figure though of shots faced Vs goals conceded doesn't look great before you get into the more complex algorithm based stats. I'm quite shocked and quite impressed with our defence last season on that basis. Watching games there were matches when it looked like we wouldn't get a shot away but I can't remember alot of games where we stopped them shooting.

[IlsonDerby]

1 hour ago, David said:

This new x (expected) stat that’s being thrown around is the worst stat in football. 

I expect better from Opta. Bin it now.

It’s on par with the level of nonsense that is Sky Sports News using video game stats to compare players.

Expected goals is very useful. It was clear under Rowett and Lampard we were over reliant on clinical finishing from Vydra/Wilson/Mount etc. 

 

I know you’ll say ‘the point is that we have clinical players’ but I’d much rather be creating chances that anyone could score rather than just one or two players. 

[Sparkle]

He was poor by his own standards, lots of soft efforts going through him and a lack dominance in his own six yard line which did not help at all 

[Day]

Just now, IlsonDerby said:

Expected goals is very useful. It was clear under Rowett and Lampard we were over reliant on clinical finishing from Vydra/Wilson/Mount etc. 

 

I know you’ll say ‘the point is that we have clinical players’ but I’d much rather be creating chances that anyone could score rather than just one or two players. 

Don’t see any use at all, most teams are reliant on one or two players putting the ball away. Lose them, you struggle. aG (actual goals) is a far more important stat and one that actually means something. 

[Woodley Ram]

2 hours ago, therealhantsram said:

Opta’s review of 2019 ranks Scott Carson 24th out of 24 Championship keepers.  ?

I knew his performance wasn’t up to previous seasons but didn’t realise just how bad he was last term.

 

https://www.optasportspro.com/static/10-create-season-review.html

 

Absolute tosh, Carson is and will remain one of the best goalkeepers in the Championship...end of..

[RoyMac5]

9 minutes ago, IlsonDerby said:

Expected goals is very useful. It was clear under Rowett and Lampard we were over reliant on clinical finishing from Vydra/Wilson/Mount etc. 

 

I know you’ll say ‘the point is that we have clinical players’ but I’d much rather be creating chances that anyone could score rather than just one or two players. 

How do you know that anyone could score them? 6 yard tap-ins, I've seen them missed!

This is why goal-scorers command such silly fees. The reason 'the stats' (whatever that means, sounds like an autonomous being rather than a man-made mathematical equation) show they scored so many is that they were our main goal threats.

[therealhantsram]

7 minutes ago, David said:

Don’t see any use at all, most teams are reliant on one or two players putting the ball away. Lose them, you struggle. aG (actual goals) is a far more important stat and one that actually means something. 

That's exactly the point of xG and why it's useful. 

Actual goals is a measure of past performance, xG is a predictor of future performance. Different stats for different purposes. 

 

[Duracell]

Think he was a victim of a change in style of play and a new weakness in our defence which exposed his flaws.

He was incredible under Rowett. He had a solid defence ahead of him and all he had to do was be a shot stopper, which he excels at.

Under Lampard, he had a higher line ahead of him and fullbacks overlapping which lead to more crosses coming in. He’s not the quickest off the mark so he’s pretty planted on his line and this is bad news for the way we play.

It’s unfortunate for him because he’s a better keeper than Roos in every department apart from the two which count the most the way we currently play - distribution and springing off the line.

[Day]

4 minutes ago, therealhantsram said:

That's exactly the point of xG and why it's useful. 

Actual goals is a measure of past performance, xG is a predictor of future performance. Different stats for different purposes. 

 

You’re using past performances to predict future performances, and if anything supporting Derby would have taught you the Championship is completely unpredictable. 

Draw with Blackburn 0-0, then beat Brentford 3-1 and travel to Bolton lose 1-0. 

Every game is different, it’s not played by robots, so many different variables which make this stat utterly pointless. 

If it had any basis to it you would be able to accurately predict results right, that would put bookies out of business, why is this not happening or is it just me losing money each week?

Opta tell you this is your chance of scoring from different positions.

Looks fair enough right?

Ok, so let’s look at the game v Brentford where Wilson and Bryson hit them on the counter, one defender. Wilson puts the ball away from < 15% chance. Great!

But wait, they go on to say

The factors taken into account when assessing the quality of a chance include:

Distance from goal

Angle of the shot

Did the chance fall at the player's feet or was it a header?

Was it a one on one?

What was the assist like? (eg long ball, cross, through ball, pull-back)

In what passage of play did it happen? (eg open play, direct free-kick, corner kick)

Has the player just beaten an opponent?

Is it a rebound?

Which leaves these nerds in their offices to come up with their own interpretation of how good the chance was. This is no different to me and 200 fans sitting here after beating Wigan and discussing was Bennett’s goal a fluke or not.

[Millenniumram]

Think you could tell how bad he was with your eyes tbf without needing the stats, but they emphasise the point further. Wouldn’t have called him the worst goalie in the league but given we didn’t face that many shots, you can see how he would be from how few clean sheets we kept. His distribution was woeful and put us in serious danger, and there were quite a few goals you’d think he could’ve maybe saved and didn’t. It was no surprise when Roos came in, and it was ultimately quite easy for him to displace him.

Really not sure what happened to Scott last season as he’s previously been one of our best goalkeepers of the modern age. I’m hoping it was a one off blip, but he looked a little lethargic all season in terms of his reflexes etc, I worry it may continue. Think those stats emphasise the point I’ve been making that a new goalkeeper is a priority for us, maybe not priority number one, but it’s something that’s gotta be found this window imo.

[1of4]

2 hours ago, therealhantsram said:

Opta’s review of 2019 ranks Scott Carson 24th out of 24 Championship keepers.  ?

I knew his performance wasn’t up to previous seasons but didn’t realise just how bad he was last term.

 

https://www.optasportspro.com/static/10-create-season-review.html

 

Another stats table that proves nothing.

Two promoted teams that have the fourth and fifth worst keepers. I don't think so.

[DarkFruitsRam7]

5 minutes ago, 1of4 said:

Another stats table that proves nothing.

Two promoted teams that have the fourth and fifth worst keepers. I don't think so.

Nyland and Villa were hopeless at the time, hence why they replaced him and Bruce. Not sure about Krul but I’ve never heard people rave about him like they rave about, for example, Randolph or Henderson.

[rammieib]

If I wanted to score against Carson last season, just shoot at his feet.....

He was extremely poor..... all this tosh of a stat does is back that opinion up.

 

[Duracell]

24 minutes ago, David said:

You’re using past performances to predict future performances, and if anything supporting Derby would have taught you the Championship is completely unpredictable. 

Draw with Blackburn 0-0, then beat Brentford 3-1 and travel to Bolton lose 1-0. 

Every game is different, it’s not played by robots, so many different variables which make this stat utterly pointless. 

If it had any basis to it you would be able to accurately predict results right, that would put bookies out of business, why is this not happening or is it just me losing money each week?

Opta tell you this is your chance of scoring from different positions.

Looks fair enough right?

Ok, so let’s look at the game v Brentford where Wilson and Bryson hit them on the counter, one defender. Wilson puts the ball away from < 15% chance. Great!

But wait, they go on to say

The factors taken into account when assessing the quality of a chance include:

Distance from goal

Angle of the shot

Did the chance fall at the player's feet or was it a header?

Was it a one on one?

What was the assist like? (eg long ball, cross, through ball, pull-back)

In what passage of play did it happen? (eg open play, direct free-kick, corner kick)

Has the player just beaten an opponent?

Is it a rebound?

Which leaves these nerds in their offices to come up with their own interpretation of how good the chance was. This is no different to me and 200 fans sitting here after beating Wigan and discussing was Bennett’s goal a fluke or not.

I think it’s important to establish the difference between using stats which tell you what will happen and stats which tell you what is most likely the happen based on what often happens in similar circumstances.

 

[Day]

Just now, Duracell said:

I think it’s important to establish the difference between using stats which tell you what will happen and stats which tell you what is most likely the happen based on what often happens in similar circumstances.

I understand this is simply predicting, but again if it’s “most likely”, why are we not using this data to take the bookies to the cleaners?

How many games were we expected not to win last season but did and went to Wembley?

2-0 down away to Norwich, were Opta able to predict the floodlight failure which saw us come away with a 4-3 win.

I have no issues with predictions, we have them on this forum, we have polls where we expect to finish but we don’t give them any weight as they are just predictions. I see no reason why we should put any weight to a likely prediction based off an actual previous outcome. 

As I said earlier, so many variables you can throw in be it weather, fitness, opposition and just having an off day which factor into the game. 

[thederbyram]

9 minutes ago, David said:

I understand this is simply predicting, but again if it’s “most likely”, why are we not using this data to take the bookies to the cleaners?

How many games were we expected not to win last season but did and went to Wembley?

2-0 down away to Norwich, were Opta able to predict the floodlight failure which saw us come away with a 4-3 win.

I have no issues with predictions, we have them on this forum, we have polls where we expect to finish but we don’t give them any weight as they are just predictions. I see no reason why we should put any weight to a likely prediction based off an actual previous outcome. 

As I said earlier, so many variables you can throw in be it weather, fitness, opposition and just having an off day which factor into the game. 

I'm with you on some of this.

 

I think a lot of it should be taken in context. Last season our xG was about midtable as we'd scored a lot of goals from outside the box. But is there a caveat to say "Derby actually have a disproportionately high number of players who can shoot effectively from range"? I'm thinking Mount, Wilson (even Lawrence) should mean that when WE shoot from outside the box, the chances of us scoring are surely higher than when Bolton shot from outside the box etc. I don't think that's factored in.

 

Quality of chance is fine but when the players are of differing quality there needs to be an adjusted value surely?

 

EDIT: on reflection I don't think the stat is bad at all but the interpretation of the stat is important. By saying "here is the xG league table" we are saying "this is how difficult each chance created is and how often they're created" not "team x has been lucky/unlucky" as a lot of people tend to interpret it as.

[IlsonDerby]

1 hour ago, RoyMac5 said:

How do you know that anyone could score them? 6 yard tap-ins, I've seen them missed!

This is why goal-scorers command such silly fees. The reason 'the stats' (whatever that means, sounds like an autonomous being rather than a man-made mathematical equation) show they scored so many is that they were our main goal threats.

 

1 hour ago, David said:

Don’t see any use at all, most teams are reliant on one or two players putting the ball away. Lose them, you struggle. aG (actual goals) is a far more important stat and one that actually means something. 

I mean if neither of you can see that a high xGoals statistic is a good indicator of us playing well then this discussion cannot go any further. It is a much more in depth statistic than merely looking at 'shots/shots on target/possession' to decide on the performance of a side and whether a result was 'fair' or whether one side was more clinical. And to say 'Actual goals' are the only stat that matters reduces our game to 1/2 moments per 90 minutes. I'm paying a lot to watch my team week in week out. For me, how we play is important. Creating chances is important. Relying on X to bang the only chance of the game for us in is exciting at the time, but always worrying that we don't have enough to sustain a challenge at the top.