Other other games (2023/24)

[Interpolic]

5 minutes ago, The Butcher said:

We were awful in that second half in Paris. The pen was a disgrace, but I don't think we deserved anything.

 

Context is important like. We matched them first half, second half we were resolutely protecting a lead. Probably retreated too early as Isak said post-match. 

 

We had played Chelsea 3 days earlier, had no viable subs to bring on and Isak was hobbling around. PSG hadn't managed to break us down before getting a penalty that was beyond dodgy. 

 

Personally I think we'd have sealed the deal second half with something resembling a squad. 

[Tsunami]

I can’t remember many, if any, great chances that PSG had, especially in the 2nd half (we missed a great chance at nil nil). They were shooting from distance and we were relatively comfortable. That xg had to be largely down to the number of “tickets” they’d bought. 
 

Anyway, come on Dortmund.

[Optimistic Nut]

Think the frustration was we didn't get to play matchdays 4 to 6 with the sort of squad that could have done something. We could have nicked the home Borussia match but we were crippled games 4-5, the PSG game would have been a miracle had we won. 

 

I reckon our squad with 3-4 key injuries would have qualified.

 

 

Edited April 17 by Optimistic Nut

[The College Dropout]

19 minutes ago, Tsunami said:

I can’t remember many, if any, great chances that PSG had, especially in the 2nd half (we missed a great chance at nil nil). They were shooting from distance and we were relatively comfortable. That xg had to be largely down to the number of “tickets” they’d bought. 
 

Anyway, come on Dortmund.

They had loads of chances. Loads. Barcola in particular. They cut us open time and again.

 

We were eventually robbed. But the defensive performance was bad like it was against Chelsea in the league cup, or Chelea or Arsenal in the league.

[SteV]

A reminder of our bench options away in Paris:

 

Dubravka

Karius

Dunmett

Hall

Huntley

Ndiweni

Parkinson

 

[Optimistic Nut]

10 minutes ago, SteV said:

A reminder of our bench options away in Paris:

 

Dubravka

Karius

Dunmett

Hall

Huntley

Ndiweni

Parkinson

 

 

Add in the context of the fixture list in general around that time. I know we got thumped at the likes of Bournemouth, Spurs and Everton but to beat Chelsea and Man Utd and draw away to PSG with the squad we had is amazing looking back. 

 

 

[Tsunami]

18 minutes ago, The College Dropout said:

They had loads of chances. Loads. Barcola in particular. They cut us open time and again.

 

We were eventually robbed. But the defensive performance was bad like it was against Chelsea in the league cup, or Chelea or Arsenal in the league.

I honestly don’t remember many clear cut chances, they undoubtedly had all of the possession and got themselves in great positions, I just didn’t feel they were going to score and we had our game plan spot on. Their XG was also massively boosted by that ridiculous penalty. That might be rose tinted but I’ve felt far more nervous in other games. Against Arsenal we looked like we had no plan and they created really good chances.
 

XG is bollocks imo btw, it’s really all about context.
 

I’ve little doubt if PSG had scored 20 mins earlier that we’d have lost and deservedly so, probably by a couple of goals.

[Groundhog63]

Hindsight anyway but, given how the season progressed with injuries it was probably a blessing our involvement in Europe ended when it did. 

Doubt we'd be in with a sniff, however fragile, of 5th.

 

 

Ps appreciate further involvement may have just been a 2 leg tie v Rennes 😂

[The College Dropout]

28 minutes ago, Tsunami said:

I honestly don’t remember many clear cut chances, they undoubtedly had all of the possession and got themselves in great positions, I just didn’t feel they were going to score and we had our game plan spot on. Their XG was also massively boosted by that ridiculous penalty. That might be rose tinted but I’ve felt far more nervous in other games. Against Arsenal we looked like we had no plan and they created really good chances.
 

XG is bollocks imo btw, it’s really all about context.
 

I’ve little doubt if PSG had scored 20 mins earlier that we’d have lost and deservedly so, probably by a couple of goals.

A penalty Xg is .8-.85. Their overall Xg was roughly 4.5-4.8. 

 

They had plenty of big chances. Sofascore says the had 8 non-penalty big chances, we had 2.  Sofascore defines a Big chance - a clear-cut scoring opportunity, like a one-on-one situation or a shot from just a few yards out.

 

Tbf Arsenal we were bad for 90 minutes. The PSG game was more similar to the Chelsea game.

 

I'm a fairly avid FPL player And I find Expected stats very useful. It nearly always reflects something true to real life, especially over several games. Eg Darwin Nunez has excellent movement and gets on the end of lots of quality chances but he’s a poor finisher. He’ll routinely underperform his Xg. Son on the other hand routinely over-performs his Xg but he sometimes goes through phases where his Xg is prohibitively low, despite the strong finishing. He doesn’t do so well against low-block defences where his Xg dramatically drops.

[christ]

3 minutes ago, The College Dropout said:

A penalty Xg is .8-.85. Their overall Xg was roughly 4.5-4.8. 

 

They had plenty of big chances. Sofascore says the had 8 non-penalty big chances, we had 2.  Sofascore defines a Big chance - a clear-cut scoring opportunity, like a one-on-one situation or a shot from just a few yards out.

 

Tbf Arsenal we were bad for 90 minutes. The PSG game was more similar to the Chelsea game.

 

I'm a fairly avid FPL player And I find Expected stats very useful. It nearly always reflects something true to real life, especially over several games. Eg Darwin Nunez has excellent movement and gets on the end of lots of quality chances but he’s a poor finisher. He’ll routinely underperform his Xg. Son on the other hand routinely over-performs his Xg but he sometimes goes through phases where his Xg is prohibitively low, despite the strong finishing. He doesn’t do so well against low-block defences where his Xg dramatically drops.

xG is nonce shit. Pack it in.

[Tsunami]

Tbf my eyes tell me an awful lot of what’s being said there. I’d rather trust my eyes than someone else’s interpretation of a good chance. I agree over a large period it might be useful but in individual games it’s got to be about the context of the game? I’m assuming Xg doesn’t make any adjustments for bogey sides or he always scores against xxx? Mad though it is, in individual games it’s a thing.

 

Anyway, I felt we were relatively comfortable against an increasingly frustrated PSG.

 

On a lighter note whatever Werner’s XG is, it should be lower.

[The Prophet]

xG is like any other stat. It's useful in context, when used alongside other stats and the eye test. But on it's own it can pretty misleading.

[The Prophet]

Also PSG created a few very good chances, but we were saved by good goalkeeping and poor finishing. The rearguard action v Milan was more convincing.

[The College Dropout]

10 minutes ago, Tsunami said:

Tbf my eyes tell me an awful lot of what’s being said there. I’d rather trust my eyes than someone else’s interpretation of a good chance. I agree over a large period it might be useful but in individual games it’s got to be about the context of the game? I’m assuming Xg doesn’t make any adjustments for bogey sides or he always scores against xxx? Mad though it is, in individual games it’s a thing.

 

Anyway, I felt we were relatively comfortable against an increasingly frustrated PSG.

 

On a lighter note whatever Werner’s XG is, it should be lower.

Yeh of course. they had 31 shots. A full 14 off target. A bunch blocked. 7 on target.

 

 Even in these shortened highlights. They had 4 or 5 great opportunities. Some really poor finishing, some great goalkeeping and great blocks.

 

Xg isn't predictive of the future. It tells a story of what has happened. 

 

It was wave after wave of attack and they were consistently generating opportunities. 

 

Xg and looking at data in general helps to reduce bias especially. IMO if we had last that 3-1 without the penalty, we couldn't complain. They sliced us open time and again and we offered little going forward for 45 minutes. But it hurts because they didn't and were gifted a penalty.

[Kid Icarus]

10 hours ago, Froggy said:

 

Why not? You're the king of "shoulds." "Should be 4 to Bournemouth." "Man United should have lost because Solanke shouldn't have been offside."

 

Well PSG absolutely battered you in that game (4.68xg - 1.40) They "should" have been 4-1 up. So should they have gone through or not? 

 

Well, putting everything aside for a moment, I think the most important thing is that you haven't let any of this get to you.

 

We went 1-0 up, it should have been 2 up. We then went backs to the wall in the 2nd half for reasons other people have mentioned and did it very well. PSG should have scored a few times.

 

Then in the 98th minute with the game over, PSG got a shockingly bad decision that handed them a draw, at that point we should have won and didn't - not because of our mistakes or PSG's brilliance - but because of a decision so bad that most neutral, unbiased viewers agreed we were robbed of a hard fought victory.

 

Context, non-mutual exclusivity and linear time. Next week: rocket science.

 

And get it right, it's "Man United should have lost by more than the 3-0 that they lost by because Solanke shouldn't have been offside." 

 

 

Edited April 17 by Kid Icarus

[Checko]

I think xG's useful, but like all stats it has its limitations. You could create nothing all game and then two defensive cockups gift you a couple of gilt edge chances and suddenly it looks like you created a bit. It's most useful when supplemented with the eye test I think.

 

Also, if someone puts a great cross across the 6 yard box and an attacker is a barleycorn away from making contact and putting it in an empty net - is that counted for xG/xA? Or is it not added to those stats as there's no contact to measure the potential shot from?

 

Also I agreed with TCD that PSG battered us for the entire second half in Paris and we probably should have conceded at least 3 in the second half.

 

 

Edited April 17 by Checko
stray '

[Kid Icarus]

52 minutes ago, christ said:

xG is nonce shit. Pack it in.

Personally think the performative hatred and deliberate misrepresentation of how people use it is the nonce shit tbh. 

 

It's just another stat that doesn't need to be taken that seriously.

 

 

Edited April 17 by Kid Icarus

[The College Dropout]

Just now, The Prophet said:

Also PSG created a few very good chances, but we were saved by good goalkeeping and poor finishing. The rearguard action v Milan was more convincing.

I think Milan was much the same but not as bad.

 

On the balance of play, we should've lost that game as comfortably as the Dortmund game.

 

I think the Xg for that game was something like AC Milan 1.8-2.2 : Newcastle 0.1-0.2.

 

8+ shots on target for Milan and 1-3 for us.

 

On the balance of play, I don't think we deserved any points away from home in the CL. 

[The College Dropout]

With Nick Pope in goal, we were that type of side to overperform our Xga. Maybe it's some psychology to it. He intimidates opposition - maybe via making a few good saves, then they try extra hard to hit the corners or blast it to beat him and the shots go astray. Maybe we make those chances a little bit harder with him due to pressure on the ball or positioning. Lots of games with Pope in goal we withstood heavy pressure. Maybe he's just that good and Dubs is just that bad. But I think Dubs started the Chelsess game actually.

 

I think Man U are that type of side. I expect the opposition to fall apart in front of goal against them. I've been waiting for the Xg/Xga stats to come to fruition all season and it doesn't (consistently). Opposition keep fluffing great chances and they keep being relatively clinical.

[christ]

17 minutes ago, Kid Icarus said:

Personally think the performative hatred and deliberate misrepresentation of how people use it is the nonce shit tbh. 

 

It's just another stat that doesn't need to be taken that seriously.

 

 

 

You are wrong. Nonce shit.

[Kid Icarus]

9 minutes ago, christ said:

You are wrong. Nonce shit.

 

Neh. Like I say, you and others take it too seriously and have now started absorbing it into your personality.

 

I'll concede that people using it as gospel, beyond what their eyes are telling them, is perversion though.

 

 

Edited April 17 by Kid Icarus

[The College Dropout]

6 minutes ago, Kid Icarus said:

 

Neh. Like I say, you and others take it too seriously and have now started absorbing it into your personality.

 

I'll concede that people using it as gospel, beyond what their eyes are telling them, is perversion though.

 

 

 

Do people have examples of when they thought Xg was wrong?

 

I think sometimes specific shots might seem out of whack with the eye test. But not whole matches typically

[Kid Icarus]

8 minutes ago, The College Dropout said:

Do people have examples of when they thought Xg was wrong?

 

I think sometimes specific shots might seem out of whack with the eye test. But not whole matches typically

I can't think of any specific examples but it's definitely happened plenty of times in my experience just with Newcastle alone.

 

It's why I don't think it's worth taking as gospel, or really investing in that much beyond looking at larger overall patterns, with a player's FBRef stats or seeing the sustainability of a team's style for example.

[Kid Icarus]

Speaking of which just because it's fresh in my mind, here's a good example actually.

 

If you watched this match you'd know 0-3 actually flattered Man United and you'd expect the xG to at least reflect the actual scoreline.

 

[The College Dropout]

34 minutes ago, Kid Icarus said:

Speaking of which just because it's fresh in my mind, here's a good example actually.

 

If you watched this match you'd know 0-3 actually flattered Man United and you'd expect the xG to at least reflect the actual scoreline.

 

I don't remember this match.

 

But looking at the stats broadly. Man U had 20 attempts. Which is a lot. But no big chances. So that 1.16 is likely an accumulation of high volume of low Xg shots.

 

Sofascore and the highlights have Bournemouth creating 4 big chances and scoring 3. They had 10 shots with 4 on target. Watching the highlights all of the goals were good chances but none were sitters. Those headers get ballooned fairly frequently which is why their individual Xg is lower than we might think. The headers weren't point blank but they were uncontested. 

 

I agree the Bournemouth big chances could cumulatively have a higher Xg. But again, good chances but not sitters or 1v1s

 

I think a wider context of the stats paints a decent picture (when watching the highlights anyway). 4 big chances to 0, Bournemouth deserved to win and Man U deserved to lose. I agree the Xg paints a closer game than what I've seen in the highlights.

 

Big chances is also a favourite stat of mine. Some models use it to mean an Xg of over 0.25, 1 in 4 is scored. But it differs model to model.

 

Thanks for sharing.

 

I remember Burnely Dyche team tended to concede a high volume of low quality chances. With Nick Pope in goal it made it look like they were lucky but really the shot quality was poor and they had a beast in goal.

 

 

Edited April 17 by The College Dropout