Metrics of Change - the Amorim Edition

[Oranges038]

I can't see that on fbref which I have found to be the most comprehensive. Maybe at season's end we'll get some article which will break it down.

I think maybe its the same re set pieces for us to be honest.

Quickly using memory:
Sociedad (open play goal for us, penalty for them)
Fulham (open play goal for us, set piece goal for them),
Ipswich (3 set piece goals for us, 2 open play goals for them)
Everton (1 set piece goal, 1 open play for them, 2 set piece goals for us)
Spurs (open play goal for them)

So last 5 games are conceded 7 with 3 being set pieces. Scored 7 with 5 being set pieces.

Tried looking it up and found this from last week, 66% of defensive errors have led to goals. True or not, I don't know, but it does feel like every little error gets punished.




Only two teams have made fewer errors leading to shots in the Premier League than Man Utd under Amorim, currently sitting at nine. But only four teams have made more errors leading to goals than Man Utd’s six under Amorim.

That’s 66% of Man Utd’s defensive errors leading to a goal, the worst record in the Premier League since Amorim took charge. Andre Onana has been the worst suspect for errors, with all three leading to goals. Across the Premier League as a whole, no player has made more errors leading to goals than Onana since Amorim took charge. The Man Utd goalkeeper is tied on three alongside Southampton defender Taylor Harwood-Bellis.

https://www.squawka.com/en/features/ruben-amorim-damning-stats-man-utd/

 

[tomaldinho1]

Understat has it.
Open play - 32.59xg for, 32.34 against
Corner/pen/free kick/set piece - 10.9 for, 14.94 against

ok awesome, thanks. Maybe end of the season we can see where this current team comes out for all this stuff versus the rest. Should be a decent amount of games by then.

 

[Teja]

Data's too noisy right now for me. Obviously, the big one is xgD/90 and that's pretty shit. I was playing around with some AI prompting and was mostly able to get it to generate a nice viz for me if folks are interested. Happy to take ideas for customization as well.



Data's from fbref.com https://fbref.com/en/squads/19538871/Manchester-United-Stats.

 

[tomaldinho1]

Updating as we hit the 20 PL game mark. As said previously will add the set piece xg in at end of season + anything else to build a block of info we can then compare with next season.

ETH	Fulham (H)	Brighton (A)	Pool (A)	Soton (A)	Palace (A)	Spurs (H)	Villa (A)	Brentford (H)	WHUM (A)	AVERAGE
xG	2.04​	1.29​	1.5​	3.03​	2.35​	0.6​	0.45​	1.77​	2.63​	1.74​
xGA	0.42​	2.14​	2.1​	1.36​	1.67​	5.41​	0.53​	0.89​	2.99​	1.94​

Amorim

Ipswich (A)

Everton (H)

Ars (A)

Forest (H)

City (A)

Bmouth (H)

Wolves (A)

Newcastle (H)

Pool (A)

Soton (H)

Brighton (H)

Fulham (A)

Palace (H)

Spurs (A)

Everton (A)

Ipswich (H)

Arsenal (H)

Leicester (A)

Forest (A)

City (H)

AVERAGE

xG

1.27​

1.76​

0.26​

1.47​

2.05​

2.29​

0.36​

1​

1.67​

3.11​

1.32​

0.2​

1.51​

1.43​

0.65​

1.4​

1.31

1.14

1.87

1.23

1.37

xGA

1.99​

0.77​

2.97​

0.65​

0.64​

1.83​

0.82​

2.24​

2.86​

1.37​

2.17​

0.56​

2.45​

1.97​

2.1​

1.04​

1.83​

1.3

0.51

0.69

1.538

Defensively we have got slightly better, offensively no change. Be interesting to see how it all shakes out. Very good little run to end the season in terms of teams close to us in the table + hopefully a late flurry of points.

 

[romufc]

Updating as we hit the 20 PL game mark. As said previously will add the set piece xg in at end of season + anything else to build a block of info we can then compare with next season.

ETH	Fulham (H)	Brighton (A)	Pool (A)	Soton (A)	Palace (A)	Spurs (H)	Villa (A)	Brentford (H)	WHUM (A)	AVERAGE
xG	2.04​	1.29​	1.5​	3.03​	2.35​	0.6​	0.45​	1.77​	2.63​	1.74​
xGA	0.42​	2.14​	2.1​	1.36​	1.67​	5.41​	0.53​	0.89​	2.99​	1.94​

Amorim

Ipswich (A)

Everton (H)

Ars (A)

Forest (H)

City (A)

Bmouth (H)

Wolves (A)

Newcastle (H)

Pool (A)

Soton (H)

Brighton (H)

Fulham (A)

Palace (H)

Spurs (A)

Everton (A)

Ipswich (H)

Arsenal (H)

Leicester (A)

Forest (A)

City (H)

AVERAGE

xG

1.27​

1.76​

0.26​

1.47​

2.05​

2.29​

0.36​

1​

1.67​

3.11​

1.32​

0.2​

1.51​

1.43​

0.65​

1.4​

1.31

1.14

1.87

1.23

1.37

xGA

1.99​

0.77​

2.97​

0.65​

0.64​

1.83​

0.82​

2.24​

2.86​

1.37​

2.17​

0.56​

2.45​

1.97​

2.1​

1.04​

1.83​

1.3

0.51

0.69

1.538

Defensively we have got slightly better, offensively no change. Be interesting to see how it all shakes out. Very good little run to end the season in terms of teams close to us in the table + hopefully a late flurry of points.

Be quite interesting how we finish the season. Generally, the form carries onto the next season, which is why I want to see us finish quite strong.

A good strong finish will give me more confidence starting next season.

 

[AjaxCunian]

Updating as we hit the 20 PL game mark. As said previously will add the set piece xg in at end of season + anything else to build a block of info we can then compare with next season.

ETH	Fulham (H)	Brighton (A)	Pool (A)	Soton (A)	Palace (A)	Spurs (H)	Villa (A)	Brentford (H)	WHUM (A)	AVERAGE
xG	2.04​	1.29​	1.5​	3.03​	2.35​	0.6​	0.45​	1.77​	2.63​	1.74​
xGA	0.42​	2.14​	2.1​	1.36​	1.67​	5.41​	0.53​	0.89​	2.99​	1.94​

Amorim

Ipswich (A)

Everton (H)

Ars (A)

Forest (H)

City (A)

Bmouth (H)

Wolves (A)

Newcastle (H)

Pool (A)

Soton (H)

Brighton (H)

Fulham (A)

Palace (H)

Spurs (A)

Everton (A)

Ipswich (H)

Arsenal (H)

Leicester (A)

Forest (A)

City (H)

AVERAGE

xG

1.27​

1.76​

0.26​

1.47​

2.05​

2.29​

0.36​

1​

1.67​

3.11​

1.32​

0.2​

1.51​

1.43​

0.65​

1.4​

1.31

1.14

1.87

1.23

1.37

xGA

1.99​

0.77​

2.97​

0.65​

0.64​

1.83​

0.82​

2.24​

2.86​

1.37​

2.17​

0.56​

2.45​

1.97​

2.1​

1.04​

1.83​

1.3

0.51

0.69

1.538

Defensively we have got slightly better, offensively no change. Be interesting to see how it all shakes out. Very good little run to end the season in terms of teams close to us in the table + hopefully a late flurry of points.

How is this offensively no change?

 

[tomaldinho1]

How is this offensively no change?

To last update, sorry did not specify

 

[FrankWhite]

Be quite interesting how we finish the season. Generally, the form carries onto the next season, which is why I want to see us finish quite strong.

A good strong finish will give me more confidence starting next season.

I think it depends on how we're fairing in the europa. Making sure we're healthy and well conditioned for those games will be prioritised.

 

[idek2346]

https:// www.reddit.com/r/reddevils/comments/1jtfwm9/chances_against_man_city/

This is why we don't see offensive change. Awful attackers.

 

[tomaldinho1]

https:// www.reddit.com/r/reddevils/comments/1jtfwm9/chances_against_man_city/

This is why we don't see offensive change. Awful attackers.

To be honest, apart from the glaring Garnacho miss from Dalot’s cross, I feel like this paints Garnacho and Hojlund in a good light. There’s a lot of potential in those two, just need more games like that and the goals will come.

 

[Teja]

Much better at attacking set pieces, which I noticed. Our assistant #1 I think is also the set piece coach? Athletic recently said something along the lines of the assistant coming forward and barking instructions as soon as there's a set play.

Much better at open play defence which is also noticeable to the eye test I'd say.

Surprised at the disrupting opp. build up stat being worse because. I thought we were pretty high up in the table for things like PPDA, opp build up pass % etc.

 

[tomaldinho1]

Much better at attacking set pieces, which I noticed. Our assistant #1 I think is also the set piece coach? Athletic recently said something along the lines of the assistant coming forward and barking instructions as soon as there's a set play.

Much better at open play defence which is also noticeable to the eye test I'd say.

Surprised at the disrupting opp. build up stat being worse because. I thought we were pretty high up in the table for things like PPDA, opp build up pass % etc.

What does 'compared to other PL teams in recent seasons' mean? Is it just weird wording for compared to other PL teams from the same season?

 

[Pogue Mahone]

What does 'compared to other PL teams in recent seasons' mean? Is it just weird wording for compared to other PL teams from the same season?

I'm guessing the centiles are more reliable when data from a few seasons is collated, rather than basing it all on a single season?

One thing's for sure, we've been pretty fecking awful both seasons. 3/6 below the 50th centile this season. 4/6 the season before. That's diabolical. More evidence to support theory about our final league position in 2023/24 making us seem less bad than the actual football we played.

 

[AjaxCunian]

Much better at attacking set pieces, which I noticed. Our assistant #1 I think is also the set piece coach? Athletic recently said something along the lines of the assistant coming forward and barking instructions as soon as there's a set play.

Much better at open play defence which is also noticeable to the eye test I'd say.

Surprised at the disrupting opp. build up stat being worse because. I thought we were pretty high up in the table for things like PPDA, opp build up pass % etc.

Improved in;
Attacking set pieces, defense in open play

Worse in;
defensive set pieces, breaking up play from the opponent


Havent improved or slightly worse in;
Attacking open play and press resistance.

I would not necessarily call this the huge improvements some say that we have made, but hopefully with Cunha, and some more signings, a fit Amad our offensive play will improve a lot.

 

[tomaldinho1]

I'm guessing the centiles are more reliable when data from a few seasons is collated, rather than basing it all on a single season?

One thing's for sure, we've been pretty fecking awful both seasons. More evidence to support theory about our final league position in 2023/24 making us seem less bad than the actual football we played.

I would like to see both each individual season + maybe a range of 2-3 combined. Each season does exist is isolation, given all teams add/remove players/managers/injuries etc. but seeing a trend with each coach is also useful.

The more I see the bolded the more I question it with every piece of data we keep seeing. It seems very simple to think xPoints say we should have been lower last season therefore it must be true, when we are seeing the xG model in general really struggle with direct sides. What I take away from the data we keep seeing is pressing is basically the lifeblood of any half decent team unless you fill the roster with incredible players (basically to punch above your weight, you must press and press well) + xG really struggles to accurately predict how well direct teams do over a season.

If ETH's 23/24 metrics are worse than 24/25 Amorim in everything but pressing and yet the results were so different, it once again calls xG into question in terms of its calculation, particularly for long range efforts. It was 'bad' from an aesthetics standpoint certainly, but maybe it was actually 'good' from a stats standpoint.

If I think of the high pressing/hard running teams this season I think of Bmouth, Arsenal, brentford, pool, spurs, Newcastle, I am not sure how @Teja made these diagrams from markstats but would be interesting to see how those teams look as well?

 

[OnceRedAlwaysRed]

'Stick to the plan' is the new mantra of desperation. I'm increasingly nervous about Amorim, there is a reason few teams these days play 3 at the back, and it really doesn't help when we don't have the players for it. The best coaches are flexible and adapt.

Exactly. This thread is like "let's try something positive in this absolutely disastrous season". Delusional.

 

[Pogue Mahone]

I would like to see both each individual season + maybe a range of 2-3 combined. Each season does exist is isolation, given all teams add/remove players/managers/injuries etc. but seeing a trend with each coach is also useful.

The more I see the bolded the more I question it with every piece of data we keep seeing. It seems very simple to think xPoints say we should have been lower last season therefore it must be true, when we are seeing the xG model in general really struggle with direct sides. What I take away from the data we keep seeing is pressing is basically the lifeblood of any half decent team unless you fill the roster with incredible players (basically to punch above your weight, you must press and press well) + xG really struggles to accurately predict how well direct teams do over a season.

If ETH's 23/24 metrics are worse than 24/25 Amorim in everything but pressing and yet the results were so different, it once again calls xG into question in terms of its calculation, particularly for long range efforts. It was 'bad' from an aesthetics standpoint certainly, but maybe it was actually 'good' from a stats standpoint.

If I think of the high pressing/hard running teams this season I think of Bmouth, Arsenal, brentford, pool, spurs, Newcastle, I am not sure how @Teja made these diagrams from markstats but would be interesting to see how those teams look as well?

I don’t think you need to get hung up on xG. There’s a load of stuff which must contribute to these metrics - that has nothing to do with xG - which paints United under ETH in 2023/24 as much closer to a bottom half team than a top half team. And that fits with the eye test. We definitely picked up a bunch of points our performances didn’t deserve (usually with a McT late goal bailing us out of the shit) Even this season ETH seemed to pick up a few undeserved points (the game against Palace where Onana pulled off the miracle double save and Eze missed an absolute sitter)

I can’t think of a single undeserved point under Amorim. And we’ve definitely got zero points from games in which the performance deserved better. Or 1 when we deserved 3 (e.g. Maguire’s late miss against Liverpool) These unlucky dropped points happened a few times. Obviously a different story in Europe, mind you.

 

[tomaldinho1]

I don’t think you need to get hung up on xG. There’s a load of stuff which must contribute to these metrics - that has nothing to do with xG - which paints United under ETH in 2023/24 as much closer to a bottom half team than a top half team. And that fits with the eye test. We definitely picked up a bunch of points our performances didn’t deserve (usually with a McT late goal bailing us out of the shit) Even this season ETH seemed to pick up a few undeserved points (the game against Palace where Onana pulled off the miracle double save and Eze missed an absolute sitter)

I can’t think of a single undeserved point under Amorim. And we’ve definitely got zero points from games in which the performance deserved better. Or 1 when we deserved 3 (e.g. Maguire’s late miss against Liverpool) These unlucky dropped points happened a few times. Obviously a different story in Europe, mind you.

It's the stat most used in relation to the statement that United were lucky last season, really we allowed a shit load of shots but then, Onana still made a number of errors, it's not like previous seasons when DDG bailed us out massively. If I think about it, what other stats are there that get referenced? It's basically the main recycled line of 'look at xPoints, we were lucky' in 23/24.

re undeserved points, likely because we won points in so few instances sadly. Plus I'd argue Soton should have killed us off and the injury/subs then papered over that game + as much as it was nice to see us be competitive Arsenal and Pool should really have beaten us in the league with the chances they created. We should have beaten City though, for sure. But generally, yes, I think higher possession less risky styles will yield less games where we seem lucky and more where we will think we have been unlucky. The real question is whether that's a good plan long term given the success we are seeing from teams who do press high and take more risk or simply sit deep, allow a lot of shots against them and counter.

 

[TheRedDevil'sAdvocate]

Just for comparison: United in 24/25 had 382 shots for and 297 against in open play, with 0.12 xG/Sh & o.15 xGA/Sh (Understat). In 23/24, we had 404 shots for and 462 shots against in open play, with 0.11 xG/Sh & o.11 xGA/Sh. We conceded significantly fewer shots, but the quality of our opponents' chances was significantly higher, too.

 

[Pogue Mahone]

It's the stat most used in relation to the statement that United were lucky last season, really we allowed a shit load of shots but then, Onana still made a number of errors, it's not like previous seasons when DDG bailed us out massively. If I think about it, what other stats are there that get referenced? It's basically the main recycled line of 'look at xPoints, we were lucky' in 23/24.

All I'm saying is that these data support the 'look at xPoints, we were lucky' narrative. Because they do. I don't think we need to get into the weeds about xG as a concept all over again.

re undeserved points, likely because we won points in so few instances sadly. Plus I'd argue Soton should have killed us off and the injury/subs then papered over that game + as much as it was nice to see us be competitive Arsenal and Pool should really have beaten us in the league with the chances they created. We should have beaten City though, for sure. But generally, yes, I think higher possession less risky styles will yield less games where we seem lucky and more where we will think we have been unlucky. The real question is whether that's a good plan long term given the success we are seeing from teams who do press high and take more risk or simply sit deep, allow a lot of shots against them and counter.

It's not a good long term plan if you're relying on Hojlund to help convert possession into chances, that's for sure! And one thing we've definitely seen this season is that there are more ways to win games than just pressing high or counter-attacking. The league winning team being the most obvious example. Slot's main impact on Liverpool has been a fairly obvious shift away from geggenpressing to more controlled possession.

 

[tomaldinho1]

All I'm saying is that these data support the 'look at xPoints, we were lucky' narrative. Because they do. I don't think we need to get into the weeds about xG as a concept all over again.



It's not a good long term plan if you're relying on Hojlund to help convert possession into chances, that's for sure! And one thing we've definitely seen this season is that there are more ways to win games than just pressing high or counter-attacking. The league winning team being the most obvious example. Slot's main impact on Liverpool has been a fairly obvious shift away from geggenpressing to more controlled possession.

But what data outside of xPoints? That's all I have ever seen referenced. Point being if that model is shown to struggle, what other sources are we forming that opinion from?

Re Pool, I feel this was something G Nev or someone said and people just didn't question, simply because Klopp was so known for pressing. Their defensive line is actually higher, their pressing stats are extremely similar (21% success in turnover leading to chance last season, 19% this season), they are still predominantly a high pressing team but gravenberch's role is different. They scored much higher for short passes last season (280 per 90 vs 246 per 90), progressive passes is almost identical, most of their possession stats are so close you would not know they had changed manager. That is why I think everyone is extremely interested in Slot this coming season, as the use of Gravenberch was great, but he really did not change a huge amount (like Amorim has done following ETH for example, it was more like Ole following Mou)

Opta have the trends for both season, they are at an almost identical place on the chart in 23/24 and 24/25 re passing sequences/speed.
23/24 - link
24/25 - link

We need a high press

 

[bosnian_red]

It's the stat most used in relation to the statement that United were lucky last season, really we allowed a shit load of shots but then, Onana still made a number of errors, it's not like previous seasons when DDG bailed us out massively. If I think about it, what other stats are there that get referenced? It's basically the main recycled line of 'look at xPoints, we were lucky' in 23/24.

re undeserved points, likely because we won points in so few instances sadly. Plus I'd argue Soton should have killed us off and the injury/subs then papered over that game + as much as it was nice to see us be competitive Arsenal and Pool should really have beaten us in the league with the chances they created. We should have beaten City though, for sure. But generally, yes, I think higher possession less risky styles will yield less games where we seem lucky and more where we will think we have been unlucky. The real question is whether that's a good plan long term given the success we are seeing from teams who do press high and take more risk or simply sit deep, allow a lot of shots against them and counter.

Onana actually had a really good league season last year. As a team we conceded 58 goals from almost 75 xG. Horrible in the CL, but he did well in the league that season. Our previous best over performance in that regard was 17/18, conceding 28 goals from 44 xG.

 

[Nogho]

What I liked so far:-

Defensive stability and solidity: when the wide CBs tuck in and the WBs drop back and the CMs sit in front if the CBs, it creates solid wall of 7 players defensive structure with the AMs and the ST ahead.

Control of the ball: though a lot of it at times is the ball being recycled between the back line and the GK, there is better control by having 3 CBs passing the ball sidaways to move the oppo players side to side, as it makes it a lot harder for the oppo to press aggressively and risk coming out, it should improve of course but the control element is there.

Emerging patterns in attack: There is still some work that needs to be done attacking wise, but I see some patterns emerging, the first goal vs Forest was a beauty, Bruno dropping deep and pulling a defender on to him that created the space for Ugarte to exploit and slot Garna in a 1v1, Hojlund ends up scoring it but it was a beauty, the winning goal against Bodo, the winning goal against Plzen, both feel like patterns rather than individual brilliance, which we still need like yesterday's game vs City, but I like the patterns and diversification of the attacks rather than being dependant on someone being brilliant on the da, still early days but promising.

What I don't like is the set peices defending, we should do a lot better there and hopefully it gets better.

I'll be honest, I actually hate 3 at the back formations, really hate and and I hate teams that we come up against that play that way, even though I watch Inter's games, not as regularly as United, but even when I watch Inter play and dominate and all I still disliked it, now when I watch us play it, it feels different don't know why.

Don't we often times end up with the ball on one of our defensive wings, and the wing back seems lost in what to do. So often times he just play the ball back to the centre back. Then this pattern can continue a couple of times until someone tries the dangerous ball straight through the middle.

When this pattern takes place it's often a ready signal for the the opposition to start their press.

This must be something that changed from pre-Amorim. I can't remember us keep getting stuck in way.

 

[Pogue Mahone]

But what data outside of xPoints? That's all I have ever seen referenced. Point being if that model is shown to struggle, what other sources are we forming that opinion from?

Re Pool, I feel this was something G Nev or someone said and people just didn't question, simply because Klopp was so known for pressing. Their defensive line is actually higher, their pressing stats are extremely similar (21% success in turnover leading to chance last season, 19% this season), they are still predominantly a high pressing team but gravenberch's role is different. They scored much higher for short passes last season (280 per 90 vs 246 per 90), progressive passes is almost identical, most of their possession stats are so close you would not know they had changed manager. That is why I think everyone is extremely interested in Slot this coming season, as the use of Gravenberch was great, but he really did not change a huge amount (like Amorim has done following ETH for example, it was more like Ole following Mou)

Opta have the trends for both season, they are at an almost identical place on the chart in 23/24 and 24/25 re passing sequences/speed.
23/24 - link
24/25 - link

We need a high press

Re your question in the opening para. The data in the graphics we’re discussing!

 

[Pogue Mahone]

Don't we often times end up with the ball on one of our defensive wings, and the wing back seems lost in what to do. So often times he just play the ball back to the centre back. Then this pattern can continue a couple of times until someone tries the dangerous ball straight through the middle.

When this pattern takes place it's often a ready signal for the the opposition to start their press.

This must be something that changed from pre-Amorim. I can't remember us keep getting stuck in way.

It happened all the time. In fact AWB was specifically identified as a pressing trigger for other teams because he was so poor on the ball and likely to cough up possession when pressed.

 

[tomaldinho1]

Onana actually had a really good league season last year. As a team we conceded 58 goals from almost 75 xG. Horrible in the CL, but he did well in the league that season. Our previous best over performance in that regard was 17/18, conceding 28 goals from 44 xG.

This is xG though? I feel like the direct teams (i.e. Onana last year) will always score high, so this season without looking it will be Spurs, Pool, Bmouth, Palace, Everton and Forest? This ties into the issue of xG, there's 100% something wrong with their calculation for longer shots.

 

[Nogho]

It happened all the time. In fact AWB was specifically identified as a pressing trigger for other teams because he was so poor on the ball and likely to cough up possession when pressed.

I see, bad memory from my part then. Now when you say it.

 

[tomaldinho1]

Re your question in the opening para. The data in the graphics we’re discussing!

Oh right - then go back to my original point then? It seems like you're getting sucked into the graphic a bit, as if all these things are somehow equal because they appear equal on a coloured wheel, when results show the pressing portion seems to be far more important than the others.

 

[Pogue Mahone]

Oh right - then go back to my original point then? It seems like you're getting sucked into the graphic a bit, as if all these things are somehow equal because they appear equal on a coloured wheel, when results show the pressing portion seems to be far more important than the others.

Dude you're making this way more complicated than it needs to be. The graphic is what it is. It chooses 6 key areas - presumably selected to give a good overview of all they key things a team needs to do - and they show us being well below the 50th centile (to put it mildly!) on 4/6 of them last season. Of the two that aren't well below the 50th centile, one is almost bang on it and one is only slightly above.

Now unless they've deliberately and cynically chosen data intended to try and paint Manchester United in a bad light (and we should probably discuss their findings in good faith here, or what's the point?) I'd say it's pretty solid evidence that our final league position flattered us. So I'm not sure why your response is getting so hung up on xG (and, now, pressing)

 

[tomaldinho1]

Dude you're making this way more complicated than it needs to be. The graphic is what it is. It chooses 6 key areas - presumably selected to give a good overview of all they key things a team needs to do - and they show us being well below the 50th centile (to put it mildly!) on 4/6 of them last season. Of the two that aren't well below the 50th centile, one is almost bang on it and one is only slightly above.

Now unless they've deliberately and cynically chosen data intended to try and paint Manchester United in a bad light (and we should probably discuss their findings in good faith here, or what's the point?) I'd say it's pretty solid evidence that our final league position flattered us. So I'm not sure why your response is getting so hung up on xG (and, now, pressing)

I've actually made it less complicated, I'm saying there is quite obviously an issue with taking a graphic like that and assuming each section is of equal importance purely because it presented in a manner that nudges you to do just that - this seems a very normal observation to make if ETH has us scoring 9 out of 100 for 'defence open play' versus Amorim who has us at 65 out of 100 and yet conceding more goals. Something is completely off with however this is being calculated, that seems a fair point to make?

 

[Teja]

What does 'compared to other PL teams in recent seasons' mean? Is it just weird wording for compared to other PL teams from the same season?

Oh I didn't spot the wording at all. I think it's an interesting idea if they're comparing across multiple seasons to get more data points. Imagine each club has one row per season with each row containing a bunch of stats and you do some percentiles.

The raw data is: https://markstats.club/teams-eng-24-25/ (some data exists for previous 5 seasons if you want to explore). Score on each of the categories is presumably derived from individual scores on these metrics. e.g., "attack open play" is probably a combination of xT, field tilt and xG. "Open play defence" is the inverse of that -- xTA, field tilt and xGA.

I think the markstats guy is pretty rigorous even if sometimes I feel he's biased against United so I don't think these are completely made up numbers. Happy to take them at face value.

I've actually made it less complicated, I'm saying there is quite obviously an issue with taking a graphic like that and assuming each section is of equal importance purely because it presented in a manner that nudges you to do just that - this seems a very normal observation to make if ETH has us scoring 9 out of 100 for 'defence open play' versus Amorim who has us at 65 out of 100 and yet conceding more goals. Something is completely off with however this is being calculated, that seems a fair point to make?

Mentioned this above but presumably the scores are based off of some weights on each metric. So if you value xTA or xGA higher than Goals Against then it results in a higher score on open play defence. The methodology isn't open so we can't really tweak them but I doubt he really has an agenda against us.

xTA for this season is 1.25 (6th) vs xTA under Ten Hag last year was 1.56 (11th). Similar numbers for field tilt and xGA. Quite a bit of improvement but you have to note the defensive crisis we went through with Ten Hag last season. I'd be curious to see the same chart when we did well under Ten Hag.

 

[tomaldinho1]

Oh I didn't spot the wording at all. I think it's an interesting idea if they're comparing across multiple seasons to get more data points. Imagine each club has one row per season and you do some percentiles.

The raw data is: https://markstats.club/teams-eng-24-25/ (some data exists for previous 5 seasons if you want to explore). Score on each of the categories is presumably derived from individual scores on these metrics. e.g., "attack open play" is probably a combination of xT, field tilt and xG. "Open play defence" is the inverse of that -- xTA, field tilt and xGA.

I think the markstats guy is pretty rigorous even if sometimes I feel he's biased against United so I don't think these are completely made up numbers. Happy to take them at face value.



Mentioned this above but presumably the scores are based off of some weights on each metric. So if you value xTA or xGA higher than Goals Against then it results in a higher score on open play defence. The methodology isn't open so we can't really tweak them but I doubt he really has an agenda against us.

xTA for this season is 1.25 (6th) vs xTA under Ten Hag last year was 1.56 (11th). Similar numbers for field tilt and xGA. Quite a bit of improvement but you have to note the defensive crisis we went through with Ten Hag last season. I'd be curious to see the same chart when we did well under Ten Hag.

Thanks, v helpful.

Yeah I'm not questioning the validity, more why someone would look at that graph and think the second one is 'better' without digging into what these stats actually mean. Case in point - quick glance and you would get a lot of posters agreeing we must have improved simply because this season 'outscores' last season overall if you added up all the numbers. Visually the immediate takeaway is the difference in size of defence open play and attacking set pieces and yet that has translated into a much worse season over a long period of games (it's not like it's a small sample size). I'm asking why that is.

 

[lex talionis]

Just for fun, I thought I'd ask ChatGPT what United should do "to improve its performances on the pitch next season". Addressing the goal scoring problem is at the top of the list, and I suspect most of us would agree. And given all the smoke around Cunha right now, it appears club management seem to agree that signing a striker is our top priority.

Key Areas for Manchester United to Improve On-Pitch Performance​

1. Address the Goalscoring Problem

• Manchester United have struggled to score consistently, reflected in a negative goal difference and underperformance in converting chances368.
• They underperformed their expected goals (xG) by a significant margin, particularly in open-play situations, and failed to convert 64% of their 'big chances'8.
• The attacking trio of Marcus Rashford, Rasmus Hojlund, and Alejandro Garnacho has shown promise, with the team scoring nearly a goal more per game and winning 42% more often when all three start together6. However, injuries and inconsistent fitness have disrupted this combination.

Action Points:

• Invest in a reliable, clinical striker to finish high-quality chances, as creativity is not the main issue—conversion is8.
• Improve squad depth to ensure key attackers are available and reduce reliance on injury-prone players6.

2. Improve Set-Piece Effectiveness

• United scored just five goals from set-pieces last season, the lowest in the Premier League, and their expected goals from set-plays was also among the worst8.
• Defensively, they conceded 10 goals from set-pieces and had one of the highest xG against from these situations8.

Action Points:

• Appoint a specialist set-piece coach to improve both attacking and defensive set-play routines.
• Consider recruiting a more commanding goalkeeper and a dominant centre-back to strengthen set-piece defense8.

3. Enhance Mentality and Game Management

• United have struggled to come from behind, managing just six comeback wins in 58 games, often failing to respond after conceding early goals2.
• They have not led for a single minute in three matches against Tottenham this season, indicating issues with starting games strongly and maintaining focus2.

Action Points:

• Focus on starting matches with greater intensity and concentration, especially against strong opponents.
• Develop leadership on the pitch to rally the team when behind and instill resilience.

4. Consistency Between Competitions

• United’s Europa League form has been far superior to their domestic performance, suggesting the squad has potential but struggles with consistency in the Premier League35.

Action Points:

• Analyze and replicate the tactical approaches and pressing intensity used in European matches for domestic games3.
• Address psychological factors and squad motivation to ensure consistent effort across all competitions.

5. Squad and Tactical Adjustments

• Injuries and players being forced out of position have contributed to poor results6.
• The team’s lack of urgency and pressing intensity has been criticized by the current manager3.

Action Points:

• Improve squad rotation and injury prevention strategies.
• Consider tactical tweaks, especially on the wings, to maximize attacking output and defensive solidity6.

Summary Table: Key Issues and Solutions​

Issue	Solution
Poor chance conversion	Sign a clinical striker, improve finishing
Set-piece weakness	Specialist coach, recruit CB/GK, new routines
Weak mentality/game management	Leadership development, fast starts
Inconsistent form	Replicate European tactics, boost motivation
Injuries/squad depth	Better rotation, targeted signings

 

[OmarUnited4ever]

Don't we often times end up with the ball on one of our defensive wings, and the wing back seems lost in what to do. So often times he just play the ball back to the centre back. Then this pattern can continue a couple of times until someone tries the dangerous ball straight through the middle.

When this pattern takes place it's often a ready signal for the the opposition to start their press.

This must be something that changed from pre-Amorim. I can't remember us keep getting stuck in way.

I think this stems from the weakness of buildig from the back and progressing the ball centrally, if you are good at progressing the ball centrally you force the opponent to be compact which then opens up the flanks to allow our WBs to push up and receive the ball, if we can't progress the ball centrally that means our opponent can spread out and press our WBs without worrying the ball getting through their lines.

 

[Nogho]

I think this stems from the weakness of buildig from the back and progressing the ball centrally, if you are good at progressing the ball centrally you force the opponent to be compact which then opens up the flanks to allow our WBs to push up and receive the ball, if we can't progress the ball centrally that means our opponent can spread out and press our WBs without worrying the ball getting through their lines.

Yeah, we often time get ”stuck”. I guess a really good ball carrier would help. It seemed like Yoro was told to stride forward, which really excelled at. Let’s hope that we can change this around next season.

 

[justsomebloke]

Just for fun, I thought I'd ask ChatGPT what United should do "to improve its performances on the pitch next season". Addressing the goal scoring problem is at the top of the list, and I suspect most of us would agree. And given all the smoke around Cunha right now, it appears club management seem to agree that signing a striker is our top priority.

Key Areas for Manchester United to Improve On-Pitch Performance​

1. Address the Goalscoring Problem

• Manchester United have struggled to score consistently, reflected in a negative goal difference and underperformance in converting chances368.
• They underperformed their expected goals (xG) by a significant margin, particularly in open-play situations, and failed to convert 64% of their 'big chances'8.
• The attacking trio of Marcus Rashford, Rasmus Hojlund, and Alejandro Garnacho has shown promise, with the team scoring nearly a goal more per game and winning 42% more often when all three start together6. However, injuries and inconsistent fitness have disrupted this combination.

Action Points:

• Invest in a reliable, clinical striker to finish high-quality chances, as creativity is not the main issue—conversion is8.
• Improve squad depth to ensure key attackers are available and reduce reliance on injury-prone players6.

2. Improve Set-Piece Effectiveness

• United scored just five goals from set-pieces last season, the lowest in the Premier League, and their expected goals from set-plays was also among the worst8.
• Defensively, they conceded 10 goals from set-pieces and had one of the highest xG against from these situations8.

Action Points:

• Appoint a specialist set-piece coach to improve both attacking and defensive set-play routines.
• Consider recruiting a more commanding goalkeeper and a dominant centre-back to strengthen set-piece defense8.

3. Enhance Mentality and Game Management

• United have struggled to come from behind, managing just six comeback wins in 58 games, often failing to respond after conceding early goals2.
• They have not led for a single minute in three matches against Tottenham this season, indicating issues with starting games strongly and maintaining focus2.

Action Points:

• Focus on starting matches with greater intensity and concentration, especially against strong opponents.
• Develop leadership on the pitch to rally the team when behind and instill resilience.

4. Consistency Between Competitions

• United’s Europa League form has been far superior to their domestic performance, suggesting the squad has potential but struggles with consistency in the Premier League35.

Action Points:

• Analyze and replicate the tactical approaches and pressing intensity used in European matches for domestic games3.
• Address psychological factors and squad motivation to ensure consistent effort across all competitions.

5. Squad and Tactical Adjustments

• Injuries and players being forced out of position have contributed to poor results6.
• The team’s lack of urgency and pressing intensity has been criticized by the current manager3.

Action Points:

• Improve squad rotation and injury prevention strategies.
• Consider tactical tweaks, especially on the wings, to maximize attacking output and defensive solidity6.

Summary Table: Key Issues and Solutions​

Issue	Solution
Poor chance conversion	Sign a clinical striker, improve finishing
Set-piece weakness	Specialist coach, recruit CB/GK, new routines
Weak mentality/game management	Leadership development, fast starts
Inconsistent form	Replicate European tactics, boost motivation
Injuries/squad depth	Better rotation, targeted signings

Comforting to see that AI remains limited to a combination of the obvious, the stupid and the fanciful.

 

[Stig]

I hope he learns from it, because he won’t progress beyond a manager in a small league otherwise

He had no desire to learn. He thinks he knows it all.

 

[Stig]

'Stick to the plan' is the new mantra of desperation. I'm increasingly nervous about Amorim, there is a reason few teams these days play 3 at the back, and it really doesn't help when we don't have the players for it. The best coaches are flexible and adapt. The problem with ETH is he had no real plan, suggestions he somehow abandoned it are wide of the mark. His system worked in Holland as he had by far the best and deepest squad, whereas in PL we didn't and its much more physical. The scary thing is we seem to be getting worse not better, sure the squad doesnt help but we should get more out of this team. RVN showed that briefly. INEOS are just more of the same, jumping from one manager to the next with no coherent philosophy.

You are either a modern day prophet or well ahead of your time. A true statement in the midst of statements that were so naively positive and should now have their posters cringing.

 

[Teja]

Small sample size but I thought it was a good time to resurrect this thread.

 

[Teja]

This is really the key thing to understand. We are a very direct team and that, for me, also means the game model is high variance. If you win the long ball or the second ball, the attack comes off very nicely and you manufactured a transition attempt by bypassing the press. If you don't win 3-4 long balls in a row, you've put yourself under sustained pressure from the opposition and potentially gave away a few corners and, given the rate at which we're going, conceded a goal.

There is no sense of progressing smoothly through the 3rds, peak Amorim ball is high risk, high reward game play.