Allsop v Roos

[Jimbo Ram]

3 minutes ago, Crewton said:

I wouldn't use Saturday's game as a stick with which to beat Allsop, or to claim that Roos is the better keeper (because I think that argument has been settled in the minds of a majority of fans), but after looking at the highlights again, I agree that his efforts to save the 2nd and 3rd goals looked a bit limp - the direction of both shots weren't far from his standing position and he either misread them or didn't react quickly enough. But neither attempts were "howlers" and he also made a couple of good saves to prevent further goals.

For the second I think you might also say Bird's attempt at a challenge was a bit limp too.....

[Jubbs]

19 minutes ago, Wolfie20 said:

No stats (no matter how 'brilliant' they are) can with any acceptable degree of certainty tell you whether or not keeper A would have saved the shots which keeper B didn't.

Jhdharrison1 Is a good account on twitter that shows how much goalkeeping analytics have come on. So I think it's fair to say, yes, Keeper A would have saved a shot that Keeper B didn't. 

Edited April 25, 2022 by Jubbs
formatting

[Gaspode]

There's no need to resort to stats to backup an argument - just use your eyes and it's pretty simple really....Allsop has a calming influence which encourages the defence to play the way the manager wants them to - Roos on the other hand has a 'wet yourself that he's going to make a balls up' influence which makes the defence nervous and prevents them from playing the way the manger wants - that means that Wazza picks Allsop ahead of Roos. If we change to a football style that means we hoof it up the pitch to a big target man (possible next season with a likely more physical type of football in League One which may negate the way Rooney wants us to play), then the judgement may swing towards shot stopping ability.....

[VulcanRam]

1 hour ago, Jubbs said:

He's the 5th worst shot stopper in the league of keepers who have played over 50% of games. He's prevented -2.7 goals, essentially meaning he's let in more goals than the average keeper would if they faced the same shots. Roos has prevented 1.4 goals.

Thanks for the stats, you may be right and I may be wrong.

But these kind of figures carry no weight with me given the unpredictable and fluctuating nature of football and the amount of variables involved in any one situation, let alone across one entire game. Fine as a broad trend indicator perhaps, but not much else. Doubt anyone would have xG'd Blackpool for 6 last week. I prefer to use my own eyes, and I've seen every single game this season; happy to make my judgement on that. 

[Jubbs]

1 minute ago, VulcanRam said:

Doubt anyone would have xG'd Blackpool for 6 last week. I prefer to use my own eyes, and I've seen every single game this season; happy to make my judgement on that. 

Blackpool had plenty of big chances, 3.90 xG is really high. 

Question for you, how often a game do you say "oooo he really should've scored that" or "what a miss" ? xG is telling you that, but in data. Any chance above 0.3xG is considered a "big chance" for example, penalties are 0.76 xG as out of the thousands and thousands of penalty data they have, 76% of them were scored, therefore is expected to go in 76/100 times. Wait until you get into pre shot and post shot xG which changes the value of it going in based on the shot path is estimated to go.

[Tamworthram]

1 hour ago, Jubbs said:

They collect hundreds of thousands of shot data, see what was saved and what wasn't. Work it from that. xG, xA, xT etc stats are brilliant and are so much better stats that just shots and possession that people are used to. It's modern and is far more representative of game state than simple possession and shot stats.

I know and I also know simple possession and shot stats don't always tell the full story. But, can anyone actually say that every other keeper has faced exactly, or almost exactly, the same shots as Allsop under almost identical circumstances? There are just far too many variables.

I'm actually a bit of a stato (a long time ago but did an A-level in Maths with statistics) but I also believe in analysis parallelisis. 

[Tyler Durden]

Am sure there a reason that Rooney hasn't selected Roos for a number of games so nuff said.

One probably is that Roos is just not suited to Rooneys kamikaze style of passing the ball around the back. 

[Seaside Ram]

2 hours ago, Jubbs said:

He should've saved both the 2nd and the 3rd. There's no way he should be getting beat by Semenyo's shot and he pulls out of the Klose header? Lets be honest here, how many people on here and on twitter would be criticising Roos if he conceded the same goals?

Exactly , the blinkered people like one named after our ex player / manager on here would've been going ballistic had that been Roos . But then you can't talk reason with folk like that ?

[IslandExile]

53 minutes ago, Jubbs said:

Blackpool had plenty of big chances, 3.90 xG is really high. 

Question for you, how often a game do you say "oooo he really should've scored that" or "what a miss" ? xG is telling you that, but in data. Any chance above 0.3xG is considered a "big chance" for example, penalties are 0.76 xG as out of the thousands and thousands of penalty data they have, 76% of them were scored, therefore is expected to go in 76/100 times. Wait until you get into pre shot and post shot xG which changes the value of it going in based on the shot path is estimated to go.

I'd like to see xG stats for individuals.....Marriott, Waghorn, Martin, Davison, Lee, Hector, Bloomer.....compared to their actual goals records.

When you've got time, like. Ta.

[Poynton ram]

Roos reminds me of Russell Hoult as he could be very good but I can remember games when he would flap and you could see the impact this had on his defenders.

[Rammy03]

2 hours ago, Jubbs said:

They collect hundreds of thousands of shot data, see what was saved and what wasn't. Work it from that. xG, xA, xT etc stats are brilliant and are so much better stats that just shots and possession that people are used to. It's modern and is far more representative of game state than simple possession and shot stats.

xT is the biggest nonsense stat I've ever heard

[Mucker1884]

21 hours ago, Bald Eagle's Barmy Army said:

Yep, I agree - and that’s why Roos has been in goal for the most part of this season 

[Jimbo Ram]

2 hours ago, Seaside Ram said:

Exactly , the blinkered people like one named after our ex player / manager on here would've been going ballistic had that been Roos . But then you can't talk reason with folk like that ?

Rubbish ?

[Jubbs]

3 hours ago, IslandExile said:

Marriott

21/22: 8 goals from 4.41 xG
20/21: 1 goal from 1.64 xG
19/20: 2 goals from 4.89 xG
18/19: 9 goals from 5.46 xG

3 hours ago, IslandExile said:

Waghorn

21/22: 1 goal from 3.79 xG
20/21: 5 goals from 4.99 xG
19/20: 12 goals from 14.82 xG
18/19: 10 goals from 6.78 xG
17/18: 16 goals from 13.58 xG

3 hours ago, IslandExile said:

Martin

21/22: 10 goals from 12.42 xG
20/21: 2 goals from 2.68 xG
19/20: 11 goals from 7.63 xG
18/19: 2 goals from 3.30 xG
17/18: 2 goals from 2.91 xG
16/17: 10 goals from 13.55 xG
15/16: 15 goals from 14.44 xG

[IslandExile]

Thanks @Jubbs.

To me, all three seem to be 'on target'.

Edited April 25, 2022 by IslandExile

[Ramifications]

Maybe it's time to give Roos another game. ?

[minesahartington]

On 24/04/2022 at 22:15, Ramifications said:

Allsop has better feet. Possibly Roos has better hands? 

Apart from being unable to catch?

[68 Guns]

It shouldn't even be up for discussion. And it's nothing to with Wembley 19. 

[VulcanRam]

15 hours ago, Jubbs said:

Blackpool had plenty of big chances, 3.90 xG is really high. 

Question for you, how often a game do you say "oooo he really should've scored that" or "what a miss" ? xG is telling you that, but in data. Any chance above 0.3xG is considered a "big chance" for example, penalties are 0.76 xG as out of the thousands and thousands of penalty data they have, 76% of them were scored, therefore is expected to go in 76/100 times. Wait until you get into pre shot and post shot xG which changes the value of it going in based on the shot path is estimated to go.

Thanks @Jubbs I can't debate on these things as I only have a rudimentary knowledge and yours far exceeds mine, but let's take Bart's miss on Saturday from a couple of yards. The whole ground goes "what a miss", yet you can't tell me stats in some way told us he was going to miss. Or is that what you're saying? Perhaps you're saying that that miss is analysed and used to predict future chances like that?

What can you tell us about that particular miss? Just helps me understand where you're coming from as I'd suggest the miss was totally unpredictable and greatly affected by the external (crowd, occasion etc) and internal (nerves, pressure) environment at that moment in time. Thanks. 

[LERam]

Like lots of players, Roos seems very much a confidence player. I think he is a little more up and down than Allsop, he could turn up one day and have a shocker and be a world beater the next....

 

Personally I prefer Allsop, he seems less erratic, and more stable. That's not to say he's not without fault!

 

Adjusting to league one life, one will have to go, and a youngster on less than 1k a week would be best as a reserve, that's my opinion anyway. Roos also needs to think about his career too, and at his age he really needs to be playing first team football