John Harrison said his shot stopping saved us over 10 goals last year compared to the average PL keeper. That's world class shot stopping in my eyes.
He described him as an awesome shot stopper in June.
Tweet
— Twitter API (@user) date
1
Season Average Rating
- Appearances
- 58
- Clean sheets
- 25
- Goals
- 0
- Assists
- 0
- Yellow cards
- 2
- Status
He described him as an awesome shot stopper in June.
It's odd that now you've found a statistical model that you like you've done a complete 180 and are now all about the stats, after spending weeks mocking people for using stats and as recently as Saturday morning saying "stats schmats" (twice IIRC). If you didn't trust the stats when you didn't like them, why do you suddenly trust them now you've found one you like? Particularly given that the very statistical model you appear to like still maintains that he was comfortably the worst performing of the goalkeepers at the top six clubs last season.
Tweet
— Twitter API (@user) date
Pseudoscience. Using models such as this as a comparative measure to say X keeper is better than Y is stupid. The model is flimsy and ignores the enormous amount of team variables that the goalkeeper position is subject to.
These stats seem to me to reflect reality a bit more. They also say he's made big improvements this season in sweeping and distribution and is above average in the PL. Again that reaffirms what I've seen.
They definitely say he was amazing at shot stopping last year.
I will always question psxg because it suggests that De Gea's form hasn't gone up and down these past 5 seasons. That is quite simply wrong. He was terrible for 3 years.
See above
So basically you like what you see. But only because you think it agrees with your view.
Bang on about +10 all you like. You also seem to ignore that other part of that where he says his chance prevention resulted in -6 goals.
It's that area which highlights the main reasons why people have an issue with his style of goalkeeping.
Like which variables? Obviously, the model is flawed but it does reflect reality to some degree.
You can check who is the exceptional shot stopper judging by "eye test" and then compare how this is reflected in those parameters. Then you can decide yourself if you take this seriously or not. De Gea had some excellent seasons, so did Lloris, Pope, now Alisson tops those. I have no interest convincing people who already made up their mind and will discredit every evidence which doesn't support their hard-rock opinion, so will definitely not waste more time on it.
Outfield performance? There's a very big difference between playing in a high-pressure Liverpool side with two assured CBs, versus the United teams we have seen in past years with little possession ability and complete discomfort across the back line. The fact that DDG's "best" seasons have coincided with the team being at their collective worst tells you all you need to know about the influence of the players in front. You can't view goalkeeping performance in isolation and compare against others.
Again the same guy says he's made big improvements in sweeping and distribution and is above average in the PL. He doesn't come for crosses I'll grant you that but keepers often cost their teams goals by coming for crosses and not claiming them.
I'm not ignoring the other stuff. I've never denied he was poor at that stuff, I just don't think it's as crucial as shot stopping. I was highlighting the shot stopping because people said he was average last season at it. He wasn't, he was awesome.
The model is flawed tbh.
He's definitely not an average PL GK. He's above that. Purely speaking from eye test.
He's definitely not an average PL GK. He's above that. Purely speaking from eye test.
Alisson was top last year and Liverpool conceded 24 goals in total. This season they already conceded 28 goals, and according to the model they were expected to concede 35 goals.
Last season De Gea was 3rd best shot stopper conceding 57 goals (5th worst defensive record in the league), but overall his best season was 17/18 when he conceded 28 while was expected to concede 38. We had 2nd best defensive record in the league.
Alisson was great last season and is great now even if Liverpool are shambles. De Gea best season saved us 10 goals and was responsible for great defensive record, while lest season he was also quite good even if we conceded more than twice as many goals as 17/18. I am not sure what is the point you're trying to make, there doesn't seem to be any pattern here.
My point is that you can't use these models to say X keeper is definitively better than Y. That's been pretty clear since the start.
And yet each time, you come back with the same numbers and stats, worded slightly differently. So I ask, what point is it that you're trying to make?
But the model which you yourself are supporting shows the importance of the other areas of the game, the mode states last season his "general shot stopping" was way ahead of the other 5 goalkeepers, yet the model has him finishing below all 5 for overall impact. Surely if nothing else, the model you are now supporting proves to you that goalkeeping isn't "80% stopping shots" which you were arguing last week?
Particularly given that the better teams concede less shots on goal generally speaking; example last season we allowed 180 shots on target against, 33 more than Arsenal, 43 more than Spurs, 65 more than Chelsea, 73 more than Liverpool & 85 more than City. And yet all of those teams goalkeepers were more effective than De Gea, again, using the model that you are supporting.
I'll accept that if you accept he's an awesome top level shot stopper and is improving in distribution and sweep
ing.I just shown you that this statement:
is not true since it doesn't apply for neither Alisson nor De Gea in terms of their best seasons.
I don't think PSxG is a good metric to compare goalkeepers, but it will tell you if any given keeper is "amazing" shotstopper what I think some people are trying to prove.
You haven't shown anything I'm afraid. And you're generally side stepping the main point I'm making.
I've happily said he's improving with his sweeping and trying to deal with crosses, I think if you read back you'll see comments praising him more than once in the last week. I've also said he's either exceptional or dreadful in terms of shot stopping and that I didn't agree with the label of "average" for that reason. I'd be intrigued to see the "model" for this season because he just isn't at the level you believe him to be, I wish he was. You seem to be under the impression that I dislike him, I don't, he's just not my style of goalkeeper. Each to their own.I'll accept that if you accept he's an awesome top level shot stopper and is improving in distribution and sweep
Can someone explain to me how any model saying that because a keeper only claimed X crosses, or swept Y times, can then attribute that back to Z additional goals conceded?
There is absolutely no way in can be accurate.
And for the benefit of bias, I'm not sure the saves metric, which has been pointed out resulted in de gea being top of the league at plus 10 can be anyway accurate either.
There is absolutely no way in can be accurate.
And for the benefit of bias, I'm not sure the saves metric, which has been pointed out resulted in de gea being top of the league at plus 10 can be anyway accurate either.
The point you made is that outfield performance impacts goalkeeper shot stopping ability. Alisson and De Gea both had great seasons when the team was doing well defensively and poor defensively. That is clear even if you don't look at stats, but the stats do reflect that. Therefore it's not an important factor in this model.
Fair enough
The situation with De Gea reminds me of the situation with Rooney when Val Gaal was in charge. He divided opinion in the same way. Fans would see Rooney miles off the top scorers list and claim we should upgrade. But then he'd have games where he'd play at a high level again and people would claim he's still our best attacker. And across all of that was the idea that people oughtn't criticise Rooney because he was a club legend and it was disrespectful.
LVG clearly had faith in Rooney or he wouldn't have played him, but there's a big gap between "good enough" and the best available. But I think LVG knew that given his club status, his respect from his teammates and his ability to turn up on the big occasion, getting rid of him would have caused such a rift that it wasn't worth doing.
With Rooney, he stuck around until Jose arrived, at which point he declined so far that it wasn't debatable any more. Indeed one of the best things Jose did at the club was to handle that situation, and he did it by simply playing Rooney and letting his football, sadly, do all the talking. No-one seriously disputed that Rooney was behind Ibra after the start to that season, and everyone knew it was time for him to leave by the end.
My guess is that DDG will continue at the club beyond this season, not because he's the best there is, there are better keepers out there for sure. Rather because he's good enough that he's not a liability, and as a result the furore of getting rid of him would outweigh the gains. I don't think DDG will go until either he decides to leave, or his form drops so low that no-one can disagree that its time to move on.
LVG clearly had faith in Rooney or he wouldn't have played him, but there's a big gap between "good enough" and the best available. But I think LVG knew that given his club status, his respect from his teammates and his ability to turn up on the big occasion, getting rid of him would have caused such a rift that it wasn't worth doing.
With Rooney, he stuck around until Jose arrived, at which point he declined so far that it wasn't debatable any more. Indeed one of the best things Jose did at the club was to handle that situation, and he did it by simply playing Rooney and letting his football, sadly, do all the talking. No-one seriously disputed that Rooney was behind Ibra after the start to that season, and everyone knew it was time for him to leave by the end.
My guess is that DDG will continue at the club beyond this season, not because he's the best there is, there are better keepers out there for sure. Rather because he's good enough that he's not a liability, and as a result the furore of getting rid of him would outweigh the gains. I don't think DDG will go until either he decides to leave, or his form drops so low that no-one can disagree that its time to move on.
The point I made is that you can't effectively compare keepers using these kinds of models due to them being inherently flawed, shallow pseudoscience.
Then it must be an amazing coincidence that season after season best and most recognized goalkeepers end up on the high end in terms of PSxG.
And I'll just hazard a guess that these keepers will generally play for the strongest, and/or best defensively organised teams as well.
Well that wording might be a bit too strong.
Let's take a look at those with 0+ PSxG over the last 3 years.
Basically De Gea gets on that list twice, in his best season 17/18 (absolute top) and last year, where we can all agree he was making some very good saves. That sounds about right considering there is wide consensus he wasn't at his best in 19-21 and a lot of people wanted him gone. So that's "eye test" reflected by stats.
Edit: btw the same model is showing that Rashford is killing it in EPL, that is what my "eye test" tells me. Not only he's accumulating xG but also converting more than expected. This metric shows me Haaland is top, but Rashy is getting close.
Last edited:
So is Kepa the 2nd best keeper this season then? Is that what this table is supposed to demonstrate?Well that wording might be a bit too strong.
Let's take a look at those with 0+ PSxG over the last 3 years.
Basically De Gea gets on that list twice, in his best season 17/18 (absolute top) and last year, where we can all agree he was making some very good saves. That sounds about right considering there is wide consensus he wasn't at his best in 19-21 and a lot of people wanted him gone. So that's "eye test" reflected by stats.
Here's how I view De Gea, all by the Mk1 eyeball, no stats involved
Makes most saves you'd expect a top-level keeper to make
Is very good with saves requiring quick reactions, especially with his feet
Makes saves that most keepers don't
Drops the odd clanger which tend to be big ones
Is very poor at commanding his area, doesn't communicate well
Is very poor on crosses but recently has improved slightly
Distribution is poor but recently has improved slightly, tends to have panic attacks
Is he the best keeper in the PL, no, is he better than average, definitely
Should he be replaced in the summer - not as a priority unless he doesn't sign a new contract in which case he's leaving anyway
Makes most saves you'd expect a top-level keeper to make
Is very good with saves requiring quick reactions, especially with his feet
Makes saves that most keepers don't
Drops the odd clanger which tend to be big ones
Is very poor at commanding his area, doesn't communicate well
Is very poor on crosses but recently has improved slightly
Distribution is poor but recently has improved slightly, tends to have panic attacks
Is he the best keeper in the PL, no, is he better than average, definitely
Should he be replaced in the summer - not as a priority unless he doesn't sign a new contract in which case he's leaving anyway
Not "keeper", no single stat will tell you that. Seems like he's been either good shot stopper or quite lucky (possible since he played less games). That's what this parameter tells you.
Would be very curious to hear from somebody who watches Chelsea regularly to say what's the story here.
https://goalkeeper.com/news-and-media/exclusives/post/shot-stopping-vs-shot-prevention
Very good article. His first example shows De Gea claiming 3 crosses against Norwich and yet all the top statistics providers said he only claimed 1 which does call into question the reliability of these stats apps.
Very good article. His first example shows De Gea claiming 3 crosses against Norwich and yet all the top statistics providers said he only claimed 1 which does call into question the reliability of these stats apps.
Shay Given saying De Gea is 100% good enough to win Utd a league when asked. Stated that the first and foremost role of a keeper is keeping the ball out of the net.
Interesting that the consensus amongst former goalkeepers is unanimously that we should not be considering moving him on, meanwhile the likes of Carl Anka, who joined him on the podcast is firmly of the belief he needs replacing. I know who's opinion carries more weight for me.
Interesting that the consensus amongst former goalkeepers is unanimously that we should not be considering moving him on, meanwhile the likes of Carl Anka, who joined him on the podcast is firmly of the belief he needs replacing. I know who's opinion carries more weight for me.
Very interesting. Quoting conclusions for those who won't get that far:
And the red bit is exactly what I've been saying about De Gea. He can't be "good", "above average", he has to be "exceptional" shot stopper to make it worth sticking to him next season.
Regarding the part about statistics providers missing some crosses, I definitely agree that is the case (I observed this myself when digging into the stats 2 years ago). Different sites have different numbers, which also are often lower than reality (not only crosses but tackles, interceptions etc).
HOWEVER I'd argue it does not play a role that much over time. This is a flaw but applicable to the model in general, so will affect numbers for all goalkeepers. But in the end if De Gea is consistently on the bottom of the table for % of intercepted crosses, season to season, and portal to portal, that means he's poor at it.
People just need to be very careful using those metrics to compare players close to each other, like places 2-7 might be very close, 8-15 and 16-20, so it's more about to which group specific player belongs rather than which position he takes on a specific metric.
Models are flawed but they will reveal extremes on both end of distribution.
We used to mock Shay Given for being exactly the type of reclusive goalkeeper De Gea has been over the past few years though, so he's not going to come out and say anything else as he likely see's De Gea as a (considerably better) version of himself.
I more have an issue with supposed experts on goalkeeping stating categorically that we can’t win the league for example with de Gea.
Meanwhile every professional/former processional keeper I’ve heard speak on it seem to be of the opinion that it would be madness to consider replacing DDG.
Have people said that? You can win the league with any kind of perceived weakness in a team if what you're doing the other end is working. For example the famous AC Milan side of 94 that won the league despite scoring only 36 goals, because their defence was exceptional and only let in 15. They flip side I guess would be our 12/13 team, that conceded 45 IIRC, and went behind in games countless times, but we had a striker that scored goals for fun and we turned them games around.
I'm intrigued how many ex goalkeepers you've heard speak on De Gea? I'm happy to discount Given for the reasons previously stated, which others have said it would be "madness" to replace him?
I think they point that was made about the pressure of being United's No.1 is considerably higher then the pressure of any other teams number 1 bar maybe Liverpool.
Even if we do replace De Gea with a keeper who can do everything, it will still be a challenge mentally that could go any which way.
Well this is lovely
Man Utd 3:1 West Ham
Important saves tonight. As usual
Two huge saves that gave us the chance to win this one.
Dave to the rescue again with his excellent chance prevention
Couldn’t do much for a goal, saved everything else, including 2 crucial saves.
- Status