Goes both ways. City play a high line and basically smother any counter attack before they start. They basically take out the first attempt at a pass. It's cynical and should result in a yellow card every time, but because it happens before the move really starts, the refs almost always let them get away with it.
United have conceded the same amount of goals in open-play as Man City in the league so far this season
- Thread starter Martial
- Start date
Goes both ways. City play a high line and basically smother any counter attack before they start. They basically take out the first attempt at a pass. It's cynical and should result in a yellow card every time, but because it happens before the move really starts, the refs almost always let them get away with it.
Look at me, I know how to post a thread, presumably.
There are lies...dam lies and then... 'statistics'.
Depends on which stats you use, when you use them and for what purpose. Conceding goals from open play tells us what?
Oh yes, we are as good as Man City... really?
Depends on which stats you use, when you use them and for what purpose. Conceding goals from open play tells us what?
Oh yes, we are as good as Man City... really?
This is irrelevant without context.
I provided context. We don't control games. You see the big teams in the league, they restrict the opposition team shooting. We are so open and allow teams far too many chances in most matches. This is so blatantly obvious and cannot be argued.
Not really. Firstly, in isolation a shot means nothing. As another poster pointed out it even included blocked shots. City conceded 22 shots to Villa when they played them…guess who had the most possession.
Broadly speaking there is an issue with control in our games but simply saying teams shoot a lot against us = no control, makes no sense.
In this sense xGA is much more useful than ‘shots’. You can clearly see Onana is operating at an extremely high level in the PL but we concede too many genuine chances. Though again # of shots means nothing - a team could have 20 por shots and never score versus 2 shots all game and win 2-0.
No they wouldn't. Some professional teams don't register 10 shots in a game.
We do seem to allow far too many chances, but we are still the 5th best defence in the league. Villa who are 4th best, have conceded only One less.
I cant wrap my head around it, but conceding/leaking goals doesnt really seem to be an issue
This is false. You can't score in football without shooting.
Plus the comparison to City is flawed; they have regressed in defense this season.
We may have the luck of the green, but it's not sustainable. The best defenses typically don't allow many shots against them.
We may have the luck of the green, but it's not sustainable. The best defenses typically don't allow many shots against them.
To be fair, thats not totally true though is it.
I knew someone would pull me up on this.
Yes a defender can decide to kick the ball into their own net. Or a ball can be hand of God'ed into the net.
But in 99% of situations, goals occur from a shot
For me our main issue up to the end of the year was scoring goals. Defensively we've been pretty good (especially with a majority of our first choice defense out for a large part of the season).
Now we have started to score we are winning games.
It's not rocket science.
Now we have started to score we are winning games.
It's not rocket science.
It shows the meaninglessness of any stat in isolation.
United have clearly had defensive issues this season. The reasons are obvious, with a new keeper and a roulette of formations due to injuries in defence/defensive midfield. The respons to that, have been a forced deployment of resources to defensive tasks, making it difficult for attacking play. Result is keeping conceded goals down, at a cost of goals scored.
The low open play conceded is interesting compared to set pieces conceded. At the same time, we concede a lot of set pieces because we struggle with shutting down open play. A lot of it comes from lack of team cohesiveness, both in moving collectively up the pitch for high pressing or fast retention after concession, and in keeping movements and passing momentum up in order to build up through high press successfully.
Neither of these issues were big issues last season (they were samll issues), and neither are explained as ‘good defence’ ‘bad defence’ ‘good strikers’ ‘bad strikers’ or even ‘good/bad tactics’.
Everton are tied with Villa in fourth for goals against.
It's variance/luck of the green right now. United and Luton have both conceded around 10.5 fewer goals than their xG. That's really not sustainable. We've also scored 5 goals fewer than our xG. Things go up and down during a season, but the underlying metrics tend to show if you're actually doing well or are on a hot run.
There is no world where conceding 20 shots per game is sustainable. It will be a problem unless we address it. Doesn't take much to realize that a hot opponent might just have a hot finishing day and we get battered as a result. Limit your opponents shots, even if pot shots (they aren't, our xG against is bottom half), and they simply didn't even take enough shots where a hot finishing day could swing the result.
In fairness - our main problem isn't the amount of goals we concede, it's the amount of goals we have scored
The set pieces were a big problem last season (for and against). Open play last season had our stats similar to Arsenal's, them being around the best set piece side in the league and us a bottom half one at both ends was a big separator.
It’s basically now similar to the Mourinho season we finished 2nd where every metric suggested we’d seriously regress at some point because we were quite lucky that year especially defensively. Seems history is repeating itself and the same pointless arguments for why we aren’t lucky because “shots don’t matter” will play out again eventually.
Or you could watch the games and see that very few of these low % shots troubled the keeper. Luton had 4 on target and I don't remember a difficult save.
Also the xG For and Against is much improved since the turn of the year. With everyone coming back from injury I'd expect that trend to continue for the rest of the season.
The fact those xG are so close does make me question the value of xG as a metric.
I think people who know football would say United created four or five "big' chances. Hojlund first goal, Garnacho one vs one, Bruno one vs one, Hojlund chance for hattrick.
We also created the Bruno chance, free shot in the box, Rashford chance at 1-0, free shot from the edge of the box and the "chance" for Hojlunds' 2nd goal - no idea how xG rates that one?
What did Luton create?
My consistent gripe with xG is how does is factor in the nuances?
So two examples...
Carlton Morris goal - free header, six yards out BUT...no pace on the ball and it's looping down after a deflection. Not commenting on whether Onana made the right or wrong decision to attempt to close down the header...but IF, all else being equal, Onana stays on his line...how does that impact xG? Does xG 'know' where Onana was? Does it 'know' there was no pace on the ball? There is a World of difference between a whipped cross that Morris simply has to nod in from that position and what actually transpired.
Likewise...the Bruno chance. He goes through on goal one vs one and takes a shot from the edge of the box at a fairly difficult angle...but does xG 'know' that he'd rounded the goalkeeper?
- Joined
- Sep 1, 2015
- Messages
- 18,322
- Supports
️
xG isn’t a problem. It’s a tool. How people understand a tool and use a tool is the problem.
The main difference is that year we were 2nd, this year we are 6th... 6th in some metrics, mid table or bottom half in others... It's been horrible. Hopefully our performances start improving and looking competent all around, rather than just some attackers doing better but an equal lack of control and unorganized defence.
We stink at set pieces this season like other seasons. Open play we are ok.
I'd ratter we had the same points as City
Looking at if a shot troubled the keeper doesn't reflect a chance quality though. Could be a great shot from 30 yards out, or could be a sitter that misses the net. XG is a stat like any other stat. Yes of course watch the games. This is a stat like any other that is used to quantify certain things. Our attacking stats have been shit this season, but we've also finished well below xG. So you'd expect it to even out, as it tends to do that. Which means if you have a bad finishing run, you'd expect to eventually have a good finishing run. That's common sense, but quantified. Likewise, if we are conceding 20 shots per game but have conceded 10 goals fewer than our xG against, you would expect to eventually get battered unless you address the issue that's leading to leaking chances. Getting away with opposition missing chances doesn't mean you defended well. It's just getting bailed out.
Is it not better to simply keep the ball, not let the opposition get in your box, rather than risking some random crosses, corners or shots against? That's what big teams do. That's what good teams do. It's nothing about Pep football or possession obsession, it's simply do the basics. Don't concede chances.
United's defense has been poor by any reason
Exactly right and you can illustrate this point further with the contrast between the 41.4 XG United has allowed, which measures the quality of chances, and the 36.2 psXG United keepers have faced (which takes XG and then adjusts for where the shot actually was targeted). That spread of 5.2 between XG and psXG is one of the highest in the league (although Liverpool is higher) and captures luck due to the opposition's shit finishing. Onana has also been good statistically allowing only 33 goals on a psXG of 36.2 (there was one own goal that doesnt get included) but the bigger factor is the opposition just finishing poorly given the chances they have had.
Another way to look at this is that United have allowed the 6th most touches in their penalty area to opponents in the league. That's not the foundation of a good defense. Let the opposition play in your penalty area often enough and bad shit will happen eventually.
Exactly right and you can illustrate this point further with the contrast between the 41.4 XG United has allowed, which measures the quality of chances, and the 36.2 psXG United keepers have faced (which takes XG and then adjusts for where the shot actually was targeted). That spread of 5.2 between XG and psXG is one of the highest in the league (although Liverpool is higher) and captures luck due to the opposition's shit finishing. Onana has also been good statistically allowing only 33 goals on a psXG of 36.2 (there was one own goal that doesnt get included) but the bigger factor is the opposition just finishing poorly given the chances they have had.
Another way to look at this is that United have allowed the 6th most touches in their penalty area to opponents in the league. That's not the foundation of a good defense. Let the opposition play in your penalty area often enough and bad shit will happen eventually.
How’s any of this luck? It’s consistent with most other seasons we have hence De Gea winning the golden gloves last season.
This is the problem with the stats. There’s deeper meaning behind them than face value but because it doesn’t paint the picture that you want it’s thrown in with luck.
Nobody is saying it’s a good defence, our back 4 is about the 4th choice back four we have at the club!
The only stats that count are final score. I get annoyed by all these stats, waste of time. Who cares how many corners, fouls, shots, possesion, red cards, yellow cards, crosses, blocks , saves, at the end of the day,did we win? did we draw? Did we lose?
I'm not saying it isn't better not to concede chances obviously.
These shots are rarely what I'd consider chances though and the xG on each one adds up, so it can look worse than the feeling you get watching the game, as the opposition are not creating much and are taking potshots which often go wide or were never on and get blocked.
We are too porous in the middle and do allow teams to attack us too often. It's quite obvious that the setup and/or personnel aren't quite right.
Some people just point to one stat over the full season and say it's bad even if that stat is showing the opposite for the run were on so far this year. It's just quite a shallow analysis.
You can watch a team and not care about anything other than the immediate result, and not care for team development and progression/growth. All the other stats are useful to provide context and to look at growth and improvement over time.
Teams can have a flukey season (United 17/18) where something like xG said we were in fact probably just the 6th best in the league. Why this matters is how projecting where the team really is. Don't fool yourself into thinking "oh we're 2nd, we should push on to the title next year, we're close". Real table might say that. Underlying numbers say we were very far away. What happened the next season? Some luck/variance went against us and we fell down to 6th which was our more real level.
That's why performance matters. When you are developing a team, your performances are what is a reliable indicator to team quality. You create chances, you restrict opposition chances, you control the game to what you want. The level at which you do that is a better indicator to how good you are. Grinding out a win despite playing like shit is great and all for the odd cup run, but it won't get you anywhere in the league.
I agree. Like I posted earlier in the thread, our expected position should be 15th based on xG expectations. That’s pretty bad. We can bury our head in the sands and say we’re 6th, but that would be poor development and learning.
Thing is that we all know we are not normally a bottom half side with ten hag. We've been horrible this season and are lucky to be 6th based on performances, and have played like a bottom half side, but there is no reason to expect us to continue playing like a bottom half side. It's just burying your head in the sand to pretend like it's been fine and not that bad.
My issue with ten hag, the way we play and things going forward is I don't think ten hag has it in him to be anything more than a top 4 competitor. The system has been horrible and while he will almost certainly adjust things (because anyone with half a brain cell will know that conceding 20 shots per game is not ok), I don't see anything to suggest he can bring us to Liverpool/Arsenal/city level. And that's what we need. We need that level of clear coaching impact. It takes time to get there, inconsistency is fine. But when you play like shit every game and start bigging up a small winning run where you didn't actually have a competent performance even once, that's when I'm going to argue it.
Once in a while we should be controlling, dominating, limiting opposition chances, and creating enough. That's not expecting a finished team, that's just expecting to occasionally have your coach actually make an impact and not make every game end to end, decided by individuals. Even Ole did that. You don't need to play super fluid football to look like a competent side. You just need to tick the basics once in a while.
People trying to pretend that stats like a bottom half xG against is not a problem is dumb. A good team just won't have that. People trying to pretend like 20 shots against per game doesn't matter is really dumb, as anybody will tell you they'd rather not concede 20 shots per game and that being 17th in the league in conceding shots is a horrendous stat. Having a while +1 goal differential at this point is horrible. It is very easy to look deeper than "6th place" (which is still shit for what it's worth), and see all the other stats that point to us performing at a horrible level. And then people will say "oh we've been better since Christmas"... Marginally yes. But in those 7 games of 5 wins, 1 draw and 1 loss, we've gained 13 xG and conceded about 11 xG, and our underlying stats put us at about 7th for this hot run. So yes, we have played a bit better since players returned from injury. But the games are still end to end and the metrics say that during this run we have been fortunate to get 5 wins, and it'll drop down. It's not sustainable performances where you are controlling every game and you can say we've turned a corner and we are deserving of this run.
It is like when DDG just joined Manchester United and the perception was that he was shite. Teams just shot on sight. Could be that Onana has that same effect now until opponents/media/people on this forum have to admit that Onana isnt that bad after all.
In my opinion xG is generally accurate but it's not infallible.
In individual games it can be off the mark as it is only as good as the samples it has to compare it to. xG can therefore be less accurate if it has kept fewer records of players, teams, matches etc.
As it based on previous data xG can therefore assign lower values to players and teams who have been performing badly but can't take into account improvements made off the field.
Over the course of a season though it will generally be quite accurate but it shouldn't be the only measure as football is not played on paper and it can't account for the human emotional side of a game.
Goals change games, scorelines with the same numbers don't mean that the games played out the same way. Any stats taken in isolation are almost irrelevant, yet there seems to be this insistence on posting xG numbers after games rather than a collection of stars together which can give you a more nuanced but not complete view of how the game was.
Corrected
The stats say it all to be honest. Isn’t he in like the top 3 this season for shot stopping?