Why does it make it useless though? Edit: It makes it flawed and perhaps I shouldn't have said that marking isn't relevant, since the overall average probably takes it into account. I have seen you using key passes in various posts. Key passes aren't useless either, but someone could easily make a point that key passes don't take into account where the shot is taken or the quality of the shot, yet key passes are still a good indication how creative a player is and that usually the most creative players have also the most key passes. How would you quantify whether a player is marked or not? My guess would be, since Lukaku's chance comes from a corner and he takes a volley, that usually in that situation the players taking that shot are tightly marked and hence miss quite a bit. And this is were the average actually helps. Because the people dismissing the 0.29 figure, somehow also conveniently ignore that Lukaku missed the chance as well, since that happens all the time, even to much better strikers than Lukaku (I also think 0.29 seems too low, just trying to explain why it might be that low).
Ole Gunnar Solskjaer | W15 D2 L4
- Thread starter Tom Van Persie
- Start date
Well yes. That's kind of what it means, isn't it? Broadly speaking. Certainly not what we do.
Useless maybe the wrong word. It makes it extremely misleading might be a better word. Key Passes doesn't pretend to rank which ones are better than others, it's just a variable, like assists and goals.
With xG however they rank each shot using a system that imo seems flawed. From my perspective, the Schlupp header was probably the best chance of the entire game. Header from a decent cross, middle of goal with entire goal to aim at. No man marked, completely open in the box. He didn't even have to stretch for it, you can see by his landing that he barely jumped. Yet that chance is deemed much lower than the Townsend chance which is a very difficult finish. The fact that its ranked actually higher than all of our goals tells you it's not very accurate imo. I also noticed that the Townsend chance is listed as being much closer to the middle of the goal than it actually was.
I think if you told a striker that you could have a shot 8 yards out from a corner on the volley completely unmarked they'd prefer it to the Townsend chance. De Geas positioning on the Townsend chance is part of the reason why it's so difficult. It's why "expected goal" is misleading, if they named it shot difficulty then i'd have less issues I guess.
No it doesn’t. The second part of my post which you decided to ignore says as such. You can chose to press and drop off depending on different phases of play or based on where you are on the pitch and the game situation.
For example, we usually press very high for the first 20 minutes or so and leave our defenders 1 v 1. I can’t remember a situation where De Gea has had to come out and sweep but I think that’s just a coincidence and will come if we keep playing this way.
Last edited:
It’s easy to comprehend why we outperform this shitty xG every time.
De Gea is much better than “average goalkeeper”, and our finishing under Ole is also much better than average finishing of thousands of Townsends who they use for calculation purposes. So we score more than xG and concede less than xGA.
What is more, our defence may sometimes even allow some shots from difficult positions to be made knowing they have de Gea
De Gea is much better than “average goalkeeper”, and our finishing under Ole is also much better than average finishing of thousands of Townsends who they use for calculation purposes. So we score more than xG and concede less than xGA.
What is more, our defence may sometimes even allow some shots from difficult positions to be made knowing they have de Gea
Tweet
— Twitter API (@user) date
Since Ole has taken over, we have the 3rd highest xG. When it comes to xG against, we are fifth. (League only)
Goals xG Goals against xG against
City 27 26.27 10 8.25
Lpool 27 22.87 8 8.05
United 26 22.43 7 13.34
Spurs 24 17.22 13 11.75
Arse 23 19.92 15 15.40
Chel 12 13.52 15 11.68
Chelsea have a game less here. It's fairly easy to interpret. City create the best xG overall and only slightly overperform. Whereas us, Liverpool, most of all Spurs and Arsenal clearly overperform according to the model. Since Ole has taken over, we have league-wide the best margin in terms of goals conceded and xG against, which is a very good indication that we have the best goalkeeper in the league and aren't lucky. It also indicates that City, Liverpool and Tottenham over the course of the season defend better than we do, which just means we can still improve in that regard.
For comparison the 17 games under Mourinho this season:
Goals xG Goals against xG against
City 48 44.98 10 13.00
Lpool 37 35.23 7 13.61
Spurs 31 29.72 16 21.46
Chel 35 31.23 14 16.44
Arse 37 27.03 23 22.08
United 29 28.22 29 26.53
What you can see, since Ole has taken over is, that we have improved in terms of finishing and that de Gea might have improved as well, since we were conceding more goals than we were expected to according to the model. Chelsea have gone from overperforming to slightly underperforming. Lloris and Alisson did concede fewer goals than they were expected to, but during that recent period Tottenham have actually conceded more than than they were expected to and Liverpool are about what the model suggests. Which could mean that both Lloris and Alisson started the season better.
For what it's worth, last season, the best teams/goalkeepers in terms of margin between goals conceded and xG against around Europe were: United/de Gea (+15.54, meaning 28 goals against and 43.54 xG against), Atletico/Oblak (+13.48), Burnley/Pope/Heaton (+13.16), Barcelona/ter Stegen (+12.62), Espanyol/Pau Lopez/Diego Lopez (+11.37), Roma/Alisson (+10.46)
Overperforming in terms of finishing compared to xG isn't an issue over the course of a season, but most of the time it suggest that isn't sustainable over several seasons. Under Conte, when Chelsea won the league, they were expected to score 61.80, but actually scored 85 goals, meaning they significantly overperformed and it wasn't sustainable, which is exactly what happened. We don't need to worry about us overperforming currently, but we need to improve in terms of creating more chances and conceding less next season, if we want to mount a title challenge, which is all the model suggests at this moment of time. What that 11tegen11 suggests is that Ole can't improve in that regard, which is a bit silly, especially if Ole gets properly backed during the summer.
Last edited:
He’s one of us. Give Ole the damn job already.
Hope he can win us the FA Cup.
You should have saved yourself some time instead of writing this. Anyone who looks at those stats and tries to interpret them is out of their mind imo. Which game since Ole joined us we won undeservedly apart from Spurs maybe (I bet the whole statistic is highly scewed by that game alone)? Herein lies the answer to you long post. In fact, arguably the only game that you can tell we got not what we deserved was Burnley, which we drew while on another day it would have been a win with the amount of chances we had.
Yep, in addition he’s a fantastic manager.
Marginally worse against Fulham according to that stat?
That’s should be a good indicator how useless it is if you actually bother watching games.xG stats are what count.At the end of the season:
Ed: Congratulations Ole you finished 3rd and won Fa Cup
Ole: Thank you, so when do I start as permanent manager??
Ed: Permanent manager?? Are you crazy have you see the xG stats you were outperformed by Fulham and Crystal Palace!! You are not getting the job
Several LFC fans have told me this.
They had two good chances at the start of the game and then another one from Babel which he should have scored. It's really not that hard to understand why they might have a good figure in that game. It still doesn't make them the better team in that game, since we actually took our chances and they didn't.Marginally worse against Fulham according to that stat?
Fulham 0 - 3 United
Fulham - 3 shots on target
United - 7 shots on target
Fulham - 2.18 xG
United - 2.11 xG
Ahuh.
Anyway, we should take this bollocks to a different topic and free this topic up of this insane dribble.
Fulham - 3 shots on target
United - 7 shots on target
Fulham - 2.18 xG
United - 2.11 xG
Ahuh.
Anyway, we should take this bollocks to a different topic and free this topic up of this insane dribble.
You cannot underestimate how the players feel about him either. Makes it very difficult to not give him the job. He deserves to be made full time manager.
Expected goals is the dumbest statistic ever created.
Ya. Ole should not be given the job based on xG.
If the account was named "BetweenTheEars" instead of "BetweenThePosts", then they would have known that it doesn't work like that
Fulham 0 - 3 United
Fulham - 3 shots on target
United - 7 shots on target
Fulham - 2.18 xG
United - 2.11 xG
Ahuh.
Anyway, we should take this bollocks to a different topic and free this topic up of this insane dribble.
0:50 and at 1:20
Babel's chance at 2:15.
I don't think that there's a whole lot wrong with xG per se. It's just a reflection of the totality of goals scored from a specific position (possibly taking into account other factors). The thing to remember is that better players score more often from those positions - and better keepers save more shots than xG would suggest. That's why they're better.
«Football analytics consultant»
Which seems very difficult for some to comprehend.
You argued previously that shooting on the volley for Lukakus chance made it significantly lower in value, yet use the first one (which is even harder given his position) as a way of pointing out it's right? Seriously!?!!? That was even graded a better chance than Lukaku's attempt against Palace
I'm done discussing this Bucky. It's utter shit.
Please please don't infest this thread with pathetic xG stats. There must be another thread on here for boring stat based chat.
I never said significantly. I was guessing why that particular chance is as low as 0.29, while you conveniently ignore that I said that it seems to low for that chance and I tried to come up with an explanation. The difference between both chances is marginal in terms of xG, one is 029, while the other is 0.37. I never said that the xG model is perfect and can't be improved upon. I am not trying to point out it is right, I am trying to explain how they come up with their numbers.You argued previously that shooting on the volley for Lukakus chance made it significantly lower in value, yet use the first one (which is even harder given his position) as a way of pointing out it's right? Seriously!?!!? That was even graded a better chance than Lukaku's attempt against Palace
I'm done discussing this Bucky. It's utter shit.
xG may be worth looking at as a small part of a bigger picture, I’ll give it that. But xG on its own doesn’t really indicate anything other than one insisting on it being a wanker who needs lashing. My two cents.
Well at this point, it would be stupid for any other managers to apply. If they got the job the scrutiny would be unbearable and they'd just be compared to how well Solskjær was doing in the little time he had, with a depleted squad, no transfer window etc. It would be like following Fergie again, you'd be mad to apply for this job now.
Correct.
Stop discussing xG
Quit the xG talk please. Use another thread to talk about stats.
Why is 'applying' being discussed here? It's not like a job offer has been posted. I can't imagine the top clubs doing anything different than approaching coaches themselves.
Funny thing is, everybody seems to agree that xG stats are useless, still they can't stop discussing it.
Last edited by a moderator:
Football is played on grass, with 11 men and a football. Not on a computer chasing statistics.
Luckily, Ole knows this.
Luckily, Ole knows this.
Just a thought: what if we skip the value of goals and let xG stats settle the games. Maybe Liverpool one day actually will win the league? Even Poch may have a shot at the title...
I wonder if Ole has been approached by any other Clubs.
We'll have to agree to disagree on the nature of a high press.
Two reasons not to give him the job:
1. xG
2. We haven´t seen how he can deal with things when the team eventually get into a bad period.
Basically: As long as he´s winning all the games, he should not be given the job.
Looking at Oles current stats since he has become caretaker manager, 29 points from 33, we are currently in top form and are playing as good as if not better than City or Liverpool. I would give Ole the Job right now.
Actually they're not useless - you just have to understand what they're telling you (which is usually something you know already). Using the Palace game as an example: we both had an xG of around 1.5. Palace scored 1 goal because they suck at getting the ball in the net and thus were worse than the expectation (think of it as an average). It could have been because de God is a better than average keeper, but we know that's not the case in this game because he had almost nothing to do. We scored 3 because we're much better than average at getting the ball in the net - again, not something that we should find surprising. Again, the keeper may have had something to do with it, although I don't think he could have done much better. So what do xG and the actual score tell us? They suck and we rule - no surprise. (Another way to think about it: Imagine we had their chances and they had ours. We'd have still scored three and they'd have still scored one - because they suck and we rule.)
Last edited: