Alarming statistics of David de Gea in our hammerings from this season so far, and last season
- Thread starter The Oracle
- Start date
I dont get why OP would feel a burning need to be dishontest about the statistics though. He wouldnt go online and make something up to fit his own narrative?
OP: Your argument directly accuses De Gea of being at fault for at least a majority of the goals conceded in these losses. Arguments like this that completely ignore shotquality, the team in front of him and all other relevant variables that contribute to a goal conceded become complely irrelevant to your argument when you try to shoehorn in an opinin that can be disproven by a 4 second google search.
The position we are allowing these shots from, how many times did players just have to tap the ball in? How many were one on ones? How many were gifts from our defenders? How many werr contested properly? They weren't Karius levels of keeping.
He's had some bad moments over the years, but when you're getting battered by these scores, everyone could have done a lot better.
He's had some bad moments over the years, but when you're getting battered by these scores, everyone could have done a lot better.
Having a go at De Gea for being culpable in Sunday’s result and score line is a bit daft. Shot stopping is all about probabilities and with small sample sizes you are bound to get variance, especially when we conceded very high quality chances. Some of the chances we conceded had a post shot xG of over 0.7, so a statistically average goalkeeper would save them less than 30% of the time.
Put another way, if you flipped a coin and got five heads in a row, that wouldn’t be that unlikely (3.125% chance), whereas if you rolled a dice and got three five times that would be much less likely (0.0128% chance).
The point being, that when we have games where we give up multiple high quality chances and the opponents finishing is on point (ie high xG and high post shot xG, which all of the results referenced in the OP would be) then there is a strong chance we end up conceding an absurd number of goals. This (mostly) isn’t the goalkeeper’s fault.
All that being said, De Gea is a below average shot stopper and has cost us around 13 goals because of shot stopping compared to the average goalkeeper over the last four and a half seasons, which is a much more reasonable sample to look at. Below is every keepers league performance in the top five leagues in that period, comparing their goals conceded versus post shot xG, normalised by how much post shot xG each has faced. De Gea is in the bottom third with everyone else you’d expect (his last truly great season was 17/18, where this metric had him first in all of Europe) and all the top keepers are at the top.
This is combined with the fact that he is poor at preventing shots by claiming crosses and sweeping, and he puts us under pressure by going long (we don’t have the forwards to be able to do this successfully) where a better keeper would play short and keep possession thus giving us more control. A goalkeeper like Maignan or Alisson would make us a much better side.
Put another way, if you flipped a coin and got five heads in a row, that wouldn’t be that unlikely (3.125% chance), whereas if you rolled a dice and got three five times that would be much less likely (0.0128% chance).
The point being, that when we have games where we give up multiple high quality chances and the opponents finishing is on point (ie high xG and high post shot xG, which all of the results referenced in the OP would be) then there is a strong chance we end up conceding an absurd number of goals. This (mostly) isn’t the goalkeeper’s fault.
All that being said, De Gea is a below average shot stopper and has cost us around 13 goals because of shot stopping compared to the average goalkeeper over the last four and a half seasons, which is a much more reasonable sample to look at. Below is every keepers league performance in the top five leagues in that period, comparing their goals conceded versus post shot xG, normalised by how much post shot xG each has faced. De Gea is in the bottom third with everyone else you’d expect (his last truly great season was 17/18, where this metric had him first in all of Europe) and all the top keepers are at the top.
This is combined with the fact that he is poor at preventing shots by claiming crosses and sweeping, and he puts us under pressure by going long (we don’t have the forwards to be able to do this successfully) where a better keeper would play short and keep possession thus giving us more control. A goalkeeper like Maignan or Alisson would make us a much better side.
Last edited:
Is this thread on here because people are tired of saying that it's Bruno's fault that we lost 7-0, so now we've picked another player solely responsible for the loss and all of our previous hammerings?
What responsibility do the backline and Casemiro get for allowing the sheer number of chances that we conceded?
What responsibility do the backline and Casemiro get for allowing the sheer number of chances that we conceded?
I absolutely agree with this, De Gea is a problem for lots of reasons, but he isn’t to blame for Sunday’s horror show, all our defensive players (bar him) made multiple errors that lead to straightforward chances, with which Liverpool were very clinical, giving him little opportunity to affect the outcome or the score line to any meaningful degree.
more excuses for varane martinez casemiro and eth.
people have even tried to pin our recent thrashing on ole. they will stoop to any level to scapegoat
people have even tried to pin our recent thrashing on ole. they will stoop to any level to scapegoat
Or is it that we should've had more thrashings but he's managed to prevent a few of them?
He does do this annoying thing when we're getting battered where he just stands rooted for the last few goals though.
He does do this annoying thing when we're getting battered where he just stands rooted for the last few goals though.
This is the crux of the matter. Liverpool's goals on Sunday all came from our players losing their heads, rushing up the pitch to try and get a goal back (at 2-0, 3-0, 4-0, 5-0 etc it didn't fecking stop and I genuinely thought it might get to double figures at one point) leaving De Gea woefully exposed. Salah's goal for the fourth for instance. The whole team played atrociously, with absolutely zero composure in the second half. Was like we were a shit 5 a side team. Sunday was such an anomaly that it's weird to think Liverpool only just managed to beat a 10 man Newcastle a couple weeks before.
As for the other games, I honestly cannot remember many of the goals but I imagine it's a similar story. Heads drop, the red mist takes over because we're getting hammered and we completely fail to shut up shop and try to knick a goal back. If the whole team has a total mare, the goalie isn't going to come out looking great regardless of how many saves he may or may not make.
And yet Mourinho became irrelevant at the top level at precisely the moment when analytics started to be used by professional football clubs. I guess they’ve all collectively lost their minds and are all wasting millions by employing data scientists to help make decisions. Maybe someone should pluck him from the obscurity of managing a Roma team that is 22 points off the league leaders to show them how it’s done.
Superficial. I don't see how De Gea could be considered at fault for any of the goals against Liverpool, by quick recollection.
Alternative to players is a lesser priority than improving on your first choice go. GK, playmaker 8, CF and RB are our four big ones either way for me at least
What was most disappointing was the gung ho approach. I can understand still trying to get back in the game at 3-0 but it was blatantly obvious we were never coming back when it got to 4-0. At that point you have to be street wise and shut up shop and take your medicine. They were pumped up, the crowd was up and our heads were gone. It made no sense to keep pushing forward and staying exposed at the back at that point.
"At fault" implies a mistake, and that's not the issue. In this case, it's simply that, given the volume and difficulty of shots he faced, an average goalkeeper would be expected to do better (judged by comparing his performance against post-shot xG)
If a striker gets 1 tricky chance in a game, you might not blame him for missing it, but if he gets 5 such chances in a game, you'd certainly criticise him for missing them all. Same principle here, none of the 7 goals he conceded were easy stops, but to concede all 7 is a poor performance.
Having a go at De Gea for being culpable in Sunday’s result and score line is a bit daft. Shot stopping is all about probabilities and with small sample sizes you are bound to get variance, especially when we conceded very high quality chances. Some of the chances we conceded had a post shot xG of over 0.7, so a statistically average goalkeeper would save them less than 30% of the time.
Put another way, if you flipped a coin and got five heads in a row, that wouldn’t be that unlikely (3.125% chance), whereas if you rolled a dice and got three five times that would be much less likely (0.0128% chance).
The point being, that when we have games where we give up multiple high quality chances and the opponents finishing is on point (ie high xG and high post shot xG, which all of the results referenced in the OP would be) then there is a strong chance we end up conceding an absurd number of goals. This (mostly) isn’t the goalkeeper’s fault.
All that being said, De Gea is a below average shot stopper and has cost us around 13 goals because of shot stopping compared to the average goalkeeper over the last four and a half seasons, which is a much more reasonable sample to look at. Below is every keepers league performance in the top five leagues in that period, comparing their goals conceded versus post shot xG, normalised by how much post shot xG each has faced. De Gea is in the bottom third with everyone else you’d expect (his last truly great season was 17/18, where this metric had him first in all of Europe) and all the top keepers are at the top.
This is combined with the fact that he is poor at preventing shots by claiming crosses and sweeping, and he puts us under pressure by going long (we don’t have the forwards to be able to do this successfully) where a better keeper would play short and keep possession thus giving us more control. A goalkeeper like Maignan or Alisson would make us a much better side.
Can we please promote him? Quality analytical posts.
I find it weird NOT to question De Gea's ability to stop shots, given that's his strongest skill.
He is definitely not better than his peers in general, even if his Save% is much better for the entire season:
The point about this below 50% saves rate is that when we are being soundly beaten, he is definitely not pulling even an average game by his own standards. I don't see anywhere in the OP any accusation that De Gea is at fault for the loss. But he certainly didn't help.
You mention shot quality. Well, the closest we have to measure shot quality is xG. And the 7 goals we conceded on Sunday came from 2.68 xG for these 7 shots and a total of 2.92 xG for the entire game.
I really don't see any solid argument in De Gea's defence other than "the entire team crumbled". And that's not a strong argument by any means.
What a load of shite.
Mourinho rallying against analytics is the “Old man yelling at clouds” personified. The entire field at the upper echelons is now dominated by it. City and Liverpool’s titles domestically? Check. Overachieving clubs like Brentford, Brighton and Leicester in the Prem? Check. The top European sides hoovering up all the young statistical outliers? Check.
These stats arent really relevant imo. Its the quality of chances conceded by the team in these games that causes it - the defence disintegrating and leaving lots of 1 on 1 chances or tap ins for the opposition means that the shots on target are likely to be goals. Likewise if we defend better and the opponents chances are 35 yard pop shots etc then the save ratio will look better.
stats are pretty meaningless when not contextualised.
or even more simply, watch some games, the guy pulls off lots of fantastic saves.
stats are pretty meaningless when not contextualised.
or even more simply, watch some games, the guy pulls off lots of fantastic saves.
Totally, it was inexplicable. A level of such severe immaturity that brutally highlights the fragility of the team under certain conditions.
Six goals in a half, having arguably been the better team in the first, is nuts that I genuinely keep forgetting it happened
I find it weird NOT to question De Gea's ability to stop shots, given that's his strongest skill.
He is definitely not better than his peers in general, even if his Save% is much better for the entire season:
The point about this below 50% saves rate is that when we are being soundly beaten, he is definitely not pulling even an average game by his own standards. I don't see anywhere in the OP any accusation that De Gea is at fault for the loss. But he certainly didn't help.
You mention shot quality. Well, the closest we have to measure shot quality is xG. And the 7 goals we conceded on Sunday came from 2.68 xG for these 7 shots and a total of 2.92 xG for the entire game.
I really don't see any solid argument in De Gea's defence other than "the entire team crumbled". And that's not a strong argument by any means.
Post shot xG is the best measure of shot quality (it’s a measure of where the shot was taken from, how hard it’s hit and where in the goal it lands) and on that measure Liverpool managed 3.92 from 8 total shots on target. This was from an xG of 2.78. Liverpool’s finishing was exceptional on the day.
It also means that the average shot faced by De Gea had a post shot xG of 0.49. For reference, the average post shot xG amongst all shots is around 0.25.
If we take out Elliott’s 0.03 long range effort in the second half they had 3.89 post shot xG for an average of 0.55 per shot. Each one is therefore unlikely to be saved in isolation and from there it’s not impossible that all of them went in. If a goalkeeper faced a hundred 0.01 shots (long range pea rollers straight at the keeper) you’d expect him to save them all. If he faced four 0.25 shots (shots hit with power from 12 yards but not in the corner, see Salah’s volley) you wouldn’t be surprised to see one or two of them go in. The total post shot xG is the same (1), but the more difficult shots to save create more opportunity for variance.
We conceded 7 because of variance. If you flipped a coin eight times and got seven heads it wouldn’t be that unusual. You cannot look at a tiny sample and berate De Gea for the scoreline. The longer term, and therefore more valid trends, already show De Gea is no longer an elite shot stopper, so there’s no need to lay this defeat at his door.
Last edited:
Yeah it really was a bizarre performance. The first half was probably one of our best away performances this season and the best we’ve played at anfield for years. And then utter capitulation in the second half.Totally, it was inexplicable. A level of such severe immaturity that brutally highlights the fragility of the team under certain conditions.
Six goals in a half, having arguably been the better team in the first, is nuts that I genuinely keep forgetting it happened
I think the big killer was the 2nd goal. We would have come out second half feeling we could get back into it. But then to concede such a poor goal straight after kick off, the players probably felt they aren’t getting anything from this, especially with the fatigue. They were probably hoping to just meander to a respectable loss. Unfortunately Liverpool were relentless.
Do you have the post shot xG for Gakpo's 2nd, the one where he kinda chipped DDG from a narorw angle, and the Firmino shot from narrow angle?
Just curious as I felt those could have been saved on a better day.
Fotmob, who use Opta’s data, have Gakpo’s second at 0.96. It was taken from within the six yard box and landed in the bottom corner. The fact that it was a chip might make it seem like he more time to react, but realistically, from that close range it’s very difficult for a keeper with his hands at hip height to save a shot that goes past him at shoulder or head height.
The Firmino one was 0.45, the shot landed much more centrally within the goal and was taken from almost exactly the same location.
Last edited:
Gakpo's second was the hardest one to save according to post-shot xG with a psxG of 0.96, which means De Gea has about 4% chance of saving it.
Using that same post-shot xG, Alisson should be diving the wrong way or something 43% of the time for this chance (post-shot xG of 0.43):
You missed the point entirely of what Jose is saying here, and I guess what the post is trying to say.
Jose used stats all the time, quite heavily times, yet his point here is anyone that uses stats and solely stats don't understand the game, they are using stats to make a point rather than their understanding and judgement of the game.
I use a saying all the time, which is 'stats without context are meaningless', which is the crux of what Jose was saying.
just bombarding someone with stats as an argument is pointless if they are not used in the correct context.
On the flip side, it's easy to use stats to anyone's advantage no matter which side of the fence.
someone put a stat on here claiming De Gea had cost United 13 goals in the last four and a half seasons, sounds horrific but is only 3ish goals per season, which really isn't that much over the course of 38 games.
these goals could have come during a time when United were 3 goals to the good as well, on which case their value diminishes by a fair bit etc etc.
You missed the point entirely of what Jose is saying here, and I guess what the post is trying to say.
Jose used stats all the time, quite heavily times, yet his point here is anyone that uses stats and solely stats don't understand the game, they are using stats to make a point rather than their understanding and judgement of the game.
I use a saying all the time, which is 'stats without context are meaningless', which is the crux of what Jose was saying.
just bombarding someone with stats as an argument is pointless if they are not used in the correct context.
On the flip side, it's easy to use stats to anyone's advantage no matter which side of the fence.
someone put a stat on here claiming De Gea had cost United 13 goals in the last four and a half seasons, sounds horrific but is only 3ish goals per season, which really isn't that much over the course of 38 games.
these goals could have come during a time when United were 3 goals to the good as well, on which case their value diminishes by a fair bit etc etc.
Who’s solely using stats to evaluate a game? The eye test shows plenty of examples of De Gea conceding easy shots, not claiming crosses, unwilling to sweep and being average at best with his feet.
Diminishing the fact that De Gea has conceded 8% more goals than expected over a large sample by saying that maybe some of those didn’t affect the outcome of a game is irrelevant when that same logic can be applied to every other goalkeeper too. It can be reversed too: how many wins turn into draws by those extra goals conceded?
You don’t need to make the same excuse for Courtois, Navas, Oblak or Alisson who have their current reputation for being elite shot stoppers on merit, is backed up by the eye test and stats, rather than past achievement.
The 13 extra goals conceded might not seem like a lot, but a truly elite shot stopper would save a similar amount versus the average goalkeeper. A 25 goal swing over a period where we have conceded around 190 is massive. That’s around five goals less per season which would absolutely have an impact on outcomes of games and the amount of points we end the season with given how few goals are scored in football.
I don’t agree with the methodology of the OP, but the conclusion that De Gea isn’t good enough is correct even if the logic used to establish it is flawed. I’ve even argued why the stats show that despite us conceding more goals than xG would suggest in this particular game, this doesn’t mean De Gea was at fault when you look at the stats in context.
De Gea is not why we conceded seven but he is a mediocre shot stopper now.
Last edited:
Drops a shit thread. fecks off without reply
Conversely the eye test for De Gea also shows a hell of a lot of brilliant elite saves.
Interesting to hear a few pundits and journalists recently claim that De Gea has yet again bailed us out of a fair few games recently.
The stats being bandied about on this thread are generally being used to back up ones already made up mind, which highlights the cherry picked nature and lack of context to which they are being presented.
Allison and Courtois have recently made absolute howlers, ironically in the same game, but that doesn't show up on shot stopping stats for example.
Not sure I agree with your conclusions here, though I do appreciate your approach. The chance of something with 0.55 probability happening 7 times from 8 is a little over 3%. Which is not impossible odds, but is certainly unusual.
It is completely pointless to apply that statistical standard to performance in a single game. Any anyway, games where the keeper has a 0% save rate aren't that rare.
All Premier League standard goalkeepers are capable of eye catching and difficult saves, that’s why there are a lot of different goalkeepers nominated for save of the month, not just the “best” ones. Suggesting that De Gea is unique in this facet doesn’t follow if you watch other teams play. This is why it’s important to use stats to broaden your scope, as well as removing the mental bias that a couple of world class saves creates.
Deferring to the opinions of journalists and pundits makes no sense to me. Ex-players, who are largely failed managers, and football journalists, are there to create engagement and entertainment, they aren’t there for critical and thoughtful analysis. If they were capable of that, they’d work for a club.
The Alisson and Courtois mistakes you are referring to were when attempting to play the ball, not from failing to stop a preventable shot. De Gea got away with one of those against Barcelona, and caused the second goal against Brentford and the first against Leeds at home due to his ball playing. Despite the fact that we are twice as likely to go long from the keeper as Liverpool, our keeper makes more mistakes playing the ball than theirs. This thread is specifically about shot stopping.
Name one statistic that I’ve presented that is out of context in this thread. Show me why they’re wrong. Until then, hand waving away stats as “taken out of context” to defend your personal biases is meaningless.
Last edited:
Perhaps you could elaborate on that? Because the fact that the industry does exactly that implies that some people find value in it.
We can’t nail the blame onto to individual performances, but those stats on DeGea are truly shocking. Be interesting to hear what our GK coach has to say about it.
De Gea has 2 out of 38 (5,26%) and Alisson has 1 out of 34 (2,94%). Some would qualify that as rare.
The 0.55 per shot comes from the ones we actually conceded from, so conceding from all 7 has a chance of around 1.5%.
Think of it this way: if you flipped a coin 7 times and got 7 heads, would you assume the coin is moody or would you keep testing it? Larger samples are required to draw valid conclusions.
My take on DDG, based only really on my eye-test, is that he'll make simple saves look very theatrical and get too much credit for some of the saves he does make.
At the same time, his reactive, passive style of goalkeeping leads him to concede goals he should save (Gakpo's 1st on Sunday)
Caveat, I am talking about post-Jose's first season...he was excellent before then and a truly world-class goalkeeper
At the same time, his reactive, passive style of goalkeeping leads him to concede goals he should save (Gakpo's 1st on Sunday)
Caveat, I am talking about post-Jose's first season...he was excellent before then and a truly world-class goalkeeper
Neither I nor the OP are berating De Gea for the score line. My (can't speak for the OP) problem with De Gea is this unfounded belief that he is good modern goalkeeper, when he is average (at a stretch) in his best quality (shot-stopping) and from poor to very poor in the rest of the important attributes any modern goalkeeper must have.
I am not singling him or any other player out for the Liverpool fiasco. That was a bad day in the office for most if not all players involved, as well as the manager.
But De Gea truly has almost all stats going against him, on top of the "eye test" (sometimes painful to have the real-time comparison with the opposition goalkeeper, like against Barcelona).
And, to top all of this, he is the best paid goalkeeper in the world, reportedly negotiating a £200k/week new contract (some reports claim that amount would be paid only if he plays, meaning with performance add-ons).
To me, that would be borderline criminal mismanagement of club resources, if given.
Especially in the context of alternatives like David Raya reportedly being available for as much as £15-20m and would likely command lesser wages.
The interesting thing is how defensive people get who use stats to bind their opinions together, and usually pick stats to maintain their own side of the coin.
Again, stats are great, if used in context and are relevant to the discussion, here they have been used to pick apart De Gea by people whom don't rate him.
Yet when other people have referred to his brilliance in certain games the stats then get changed to something else to demean his performances.
It's never once been stated that De Gea is unique in his shot stopping abilities, moreso that he is still exceptionally talented on this department.
I'm not a fan of pundits, less so football journalism, it's often created to sell a product and so is often opinionated and controversial.
But I find it interesting when multiple outlets run with a similar theme. In this case that De Gea has been saving United via saves recently.
It shows that it's something that's been happening, and generally when it's picking up on a player it's either because a)they aren't performing or b) they are performing above their norm or an outlier.
Also interesting to note that on average we give away a lot more shots on target than the likes of City, Arsenal, Liverpool and even Chelsea.
Also we concede a lot more set pieces than City and Arsenal.
And more goals via counter attacks also.
These stats can be used to piece together a story that actually our back line has been pretty average most of the season and due to our nature of playing we are susceptible to a quick counter, context is needed because apparently dropping Maguire has sorted our defence out and De Gea has more clean sheets this season than last.
Also they could be used like this:
A keeper who is facing a substantial amount more shots per game will inherently concede more goals due to a) having to perform at a higher level more often and b) having more split second decisions to make,
Also they can be used to highlight the fact that we give away nearly double the amount of shots in second halves of games than the first half.
Again, concentration for a keeper is imperative, but if he is facing double the amount of shots later on in the game, that adds another element of context to the stat.
Basically all I'm getting here, and it's a waffle, but stats can be attributed to what ever you want, but you can manipulate them to suit any narrative you'd like.
The issue with stats and football fans is still human opinion.