Statistical Domination

[hogesar 9,618]

14 hours ago, It's Character Forming said:

Interesting table @hogesar thanks

 

I thought during Farke's first promotion season, our xG for was not as good as the actual goals we scored, but our xG against was actually better than the number of goals we conceded.  I.e. we scored more than xG predicted but also conceded more than xGA predicted ?  I remember a poster called westcoastcanary who was a big fan of stats, but his attitude IIRC was that we were lucky with our goals scored, but the fact that we conceded more than xGA predicted was because our defence was not very good....

 

As I recall it during that season our xG for improved through the season so the gap with our actual goals narrowed, but our xG against and actual goals against were always poor.  Which never boded well for what happened after promotion.

I think first promotion you may be right. Second promotion I'm 90% confident we had conceded less than our xGA predicted by some margin. 

I do think, considering we've not had a recognised DM for many matches, we've looked pretty solid defensively, and after all we've conceded something like the 3rd least in the division. We'll need to continue that throughout the season, which a DM should hopefully help with. I also think conversely a DM might help the attacking players have a bit more freedom.

[wcorkcanary 4,329]

36 minutes ago, hogesar said:

I also think conversely a DM might help the attacking players have a bit more freedom

Well also see a more attacking side to Maxs game once he knows the space he's vacating is covered by a competent  DM.

[shefcanary 2,385]

42 minutes ago, hogesar said:

 I also think conversely a DM might help the attacking players have a bit more freedom.

No "conversely" about it. I attacking midfielders are currently constrained by having to do too much defensive work, which patently they are not really up to. Get in the right DM, even our Xg will increase as well as our Xga reduce.

[Badger 2,405]

On 20/09/2022 at 17:15, PockthorpePete said:

Subjective belief ?  From the person whose only basis for this is belief. Most City fans would have us to win at Blackpool. And if we don't, we can claim that not every prediction is accurate. Just as I could predict the dog in trap one will win every race. And when it doesn't, I claim that not every prediction was intended to be perfect. To compare this Tom Foolery with weather forecasting is absurd. The weather forecast is accurate. This stuff is not. Nothing more than educated guesswork. Guesswork, full stop.

It's based on probability so you need to recognise the limitation of the "prediction." If we have a 70% probability of beating Blackpool it is the most likely result, but 3 times out of 10 we will lose to or draw. Over the course of a season the unusual/ unexpected results tend to even themselves out (but not altogether).

To take your weather forecast comparison it is mostly accurate but the more precise the requirements the less accurate it will become. It can predict with a high degree of probability (say 90%) that it will be both cloudy and sunny in Norfolk tomorrow but would find it harder to give the same degree of confidence as to whether Aylsham will be cloudy at 11.45 pm.

[Badger 2,405]

56 minutes ago, hogesar said:

I think first promotion you may be right. Second promotion I'm 90% confident we had conceded less than our xGA predicted by some margin. 

I do think, considering we've not had a recognised DM for many matches, we've looked pretty solid defensively, and after all we've conceded something like the 3rd least in the division. We'll need to continue that throughout the season, which a DM should hopefully help with. I also think conversely a DM might help the attacking players have a bit more freedom.

Our XGA this season so far is 1.23. In 20-21 it was 1.38 which is significant progress. We were "mid table" defensively (on chances created against us) as were Watford. Brentford, who obviously stayed up were significantly better at 1.17 (the second lowest in the division). 

 

[Mengo 821]

1 hour ago, shefcanary said:

No "conversely" about it. I attacking midfielders are currently constrained by having to do too much defensive work, which patently they are not really up to. Get in the right DM, even our Xg will increase as well as our Xga reduce.

How long realistically has it taken to address the glaring void of DM . Negligence on behalf of certain parties,  I would say. A lot of pressure on Hayden, if he becomes fit. Can't quite believe after all this time this crucial position has not been filled 😕 and realistically without any backup. Yes the creative midfielders need to be doing the job that they are more comfortable with .

Edited September 22, 2022 by Mengo

[Petriix 2,834]

25 minutes ago, Mengo said:

How long realistically has it taken to address the glaring void of DM . Negligence on behalf of certain parties,  I would say. A lot of pressure on Hayden, if he becomes fit. Can't quite believe after all this time this crucial position has not been filled 😕 and realistically without any backup. Yes the creative midfielders need to be doing the job that they are more comfortable with .

Or, more short term, recognising that we don't have a fit DM and therefore stop trying to play a single pivot; which is beyond the skill set of any of our currently available midfielders.

Sharing the job between two of them would cover a multitude of sins and completely unleash the third midfielder to attack.

I vaguely remember the previous manager going through a long process of trying different midfield options before arriving at a double pivot. This seemed to coincide with significant success on the pitch, at least at Championship level.

It's a shame we have to go through the same mistakes all over again... 

[hogesar 9,618]

16 minutes ago, Petriix said:

Or, more short term, recognising that we don't have a fit DM and therefore stop trying to play a single pivot; which is beyond the skill set of any of our currently available midfielders.

Sharing the job between two of them would cover a multitude of sins and completely unleash the third midfielder to attack.

I vaguely remember the previous manager going through a long process of trying different midfield options before arriving at a double pivot. This seemed to coincide with significant success on the pitch, at least at Championship level.

It's a shame we have to go through the same mistakes all over again... 

I mean, we've got more points on the board than at the same time in either of our last title wins 🙂

 

[Petriix 2,834]

36 minutes ago, hogesar said:

I mean, we've got more points on the board than at the same time in either of our last title wins 🙂

 

If that's still the case at the end of the season then I'll create a thread called 'Smithball: the impossible dream made real' and don a Dean Smith mask for our first home game in the Premier League next season.

[It's Character Forming 1,160]

5 hours ago, Petriix said:

If that's still the case at the end of the season then I'll create a thread called 'Smithball: the impossible dream made real' and don a Dean Smith mask for our first home game in the Premier League next season.

Lol not exactly offering to present MOTD in your underpants 😂

[hogesar 9,618]

Blackpool added. 1.82 xG away from home is a "very good" figure when you look at the averages for the division.

[canarydan23 4,060]

Not surprising, that was a game where we were worthy of the three points, something we've not been able to say about every win.

The start was a bit ropey, but the 25 minutes after the goal was probably the best phase of play we've enjoyed this season, how we didn't add at least one other in that period I don't know.

[hogesar 9,618]

On 14/08/2022 at 12:17, priceyrice said:

Statistically, we have 1 point out of 9!

We must be one of the only clubs around were a large chunk of our fan base is happy to accept such a shower of ****

 

On 14/08/2022 at 12:19, cambridgeshire canary said:

I mean its nice and good dominating games but does it mean anything when you get 0 points anyway?

 

On 14/08/2022 at 12:25, TIL 1010 said:

Statistically we have not won a game this season and only one since the turn of the year. What does that tell you ?

So, it is still early doors but these posts from when I first started talking about xG are good examples of why clubs pay attention to this data, and why fans naturally react emotionally first without real substance.

For some fans, we were awful, Smith was awful and I was a happy clapper for daring to suggest we weren't getting the results our performances merited. I said if we kept up our general xG trend the results would change and we'd see that reflected as the season continued.

So far, that has been the case. Our xG for has generally been around the same (yesterdays was our highest, but only by 0.02). 

[hogesar 9,618]

For reference, over the weekend only Watford had a higher xG than ours in normal chances created. West Brom's was marginally higher but included a penalty which doesnt necessarily indicate the quality of chance creation. 

[hogesar 9,618]

Just a couple more bits, which isn't completely up to date. Have to feel sorry for Bruce, who's team are doing "most things right"

 

[shefcanary 2,385]

21 minutes ago, hogesar said:

Just a couple more bits, which isn't completely up to date. Have to feel sorry for Bruce, who's team are doing "most things right"

 

Wilder must be pissed off, he achieved great stats but shows his team were not putting the chances away! He was probably one player short of a good season.

[hogesar 9,618]

1 hour ago, shefcanary said:

Wilder must be pissed off, he achieved great stats but shows his team were not putting the chances away! He was probably one player short of a good season.

Yes. Although one good player short is interesting. You always think there's strikers that are 'more clinical' than others, and I certainly believe that's true.

However, the xG model throws up something interesting. There is only one player in world football since xG records were kept who regularly exceeded the amount of goals the chances they had 'were expected' to score. That's Messi. 

[Fr. Chewy Louie 137]

2 minutes ago, hogesar said:

Yes. Although one good player short is interesting. You always think there's strikers that are 'more clinical' than others, and I certainly believe that's true.

However, the xG model throws up something interesting. There is only one player in world football since xG records were kept who regularly exceeded the amount of goals the chances they had 'were expected' to score. That's Messi. 

That's interesting, does the "chance" start from when the player first gets the ball? If he gets the ball in his own half, then at that point the xG is presumably miniscule, but if he then dribbles himself into a one on one it would be fairly high. Messi scores more of that type of goal than tap ins, so would that account for it?

[hogesar 9,618]

1 hour ago, Fr. Chewy Louie said:

That's interesting, does the "chance" start from when the player first gets the ball? If he gets the ball in his own half, then at that point the xG is presumably miniscule, but if he then dribbles himself into a one on one it would be fairly high. Messi scores more of that type of goal than tap ins, so would that account for it?

I believe it's based around the point the player takes the shot. That's when xG is valued.

[Hank shoots Skyler 2,094]

On 03/10/2022 at 12:44, hogesar said:

 

 

So, it is still early doors but these posts from when I first started talking about xG are good examples of why clubs pay attention to this data, and why fans naturally react emotionally first without real substance.

For some fans, we were awful, Smith was awful and I was a happy clapper for daring to suggest we weren't getting the results our performances merited. I said if we kept up our general xG trend the results would change and we'd see that reflected as the season continued.

So far, that has been the case. Our xG for has generally been around the same (yesterdays was our highest, but only by 0.02). 

Although there's probably a West Brom supporting Hogesar who's made a similar thread on their forum, still waiting to be proven right...

[Canaries north 151]

On 03/10/2022 at 12:44, hogesar said:

 

 

So, it is still early doors but these posts from when I first started talking about xG are good examples of why clubs pay attention to this data, and why fans naturally react emotionally first without real substance.

For some fans, we were awful, Smith was awful and I was a happy clapper for daring to suggest we weren't getting the results our performances merited. I said if we kept up our general xG trend the results would change and we'd see that reflected as the season continued.

So far, that has been the case. Our xG for has generally been around the same (yesterdays was our highest, but only by 0.02). 

Yes you can use this to prove your point but take a look at West Bromwich to see how these statistics can also be misleading. If you have Pukki and serve him he might not exceed XG but will be closer than almost all strikers in the division. Some people mention shots taken etc and that is a stupid stat and XG tries to make a better judgement of that and does but it's not perfect. XG shows the chances you are creating but it's the quality of the striker that gets you points. 

[Mason 47 1,373]

Boro I believe were very close to or even top of the xG table for a long time, like West Brom they've struggled. We've never been saying xG was an on-off switch for forecasting greatness, just that at a time when a lot of people wanted the baby out with the bathwater it was worth keeping a cool head and carrying on. In our case, it's worked out.

Actually, if anything the situation has flipped in that absurd football way. Where performances should have dictated more points than we had- which means nothing as its a results game- recently we've not performed as well but got the points anyway- which means nothing as the performances aren't there.

I've held for a long time, that this month- and more intensely, Watford, Sheffield Utd and Burnley away in the space of 10 days- is when we can really analyse the whole.

Can't wait 🙈

[Robert N. LiM 4,241]

this continues to be a really interesting thread and shows how stats can provide a good framework for discussion of how your team is playing. Cheers, @hogesar

[wcorkcanary 4,329]

I love stats. Though having short term memory loss as me mum dropped me on my head when i was a Baby, i cant remember enough to argue a point. For me there are ' could  chances' ie half chances , and 'Should Chances ' ..the ones that you say how the ferk ? If you dont score.

The Messi thing is very interesting, i suppose its evidence that he's special . Extra special when you consider that  hes small and not particularly  fast , just an outstanding footballer .

Erling Haalands stats after 15 or so years at the top will be interesting too.  

[hogesar 9,618]

1 hour ago, Hank shoots Skyler said:

Although there's probably a West Brom supporting Hogesar who's made a similar thread on their forum, still waiting to be proven right...

Hey. Let's not just assume there's copy-and-paste versions of me, 'eh? 😉

On a more serious note - Bruce must be wondering what more he can do as a manager. Statistically he's set them up to create enough chances to win games and not concede too many. Of course, and where stats fail to tell a story, is how many more games WBA can perform the same, as confidence must take a hit.

[repman 634]

One thing that should be taken into consideration with both WBA and Boro's xG numbers is the fact that they have (probably) been behind in a lot of games this year. Game state can be a big factor in xG. As a match progresses, a team hanging on to a lead is going to drop deeper, probably give up more shots while having less themselves. Contrast this to us currently, when we have been leading going into the last part of the game, it's more likely we are the ones under pressure. Saying that however, most top sides will probably have a 2+ goal cushion and take the sting out of the game through possession, Man City being the masters at it.

[Robert N. LiM 4,241]

1 hour ago, Mason 47 said:

I've held for a long time, that this month- and more intensely, Watford, Sheffield Utd and Burnley away in the space of 10 days- is when we can really analyse the whole.

 

this is true, though it's also the case that putting away the 'lesser' teams is just as important, as we proved in 20-21 by winning the league at a canter despite losing home and away to Watford and Bournemouth

[Robert N. LiM 4,241]

1 minute ago, repman said:

One thing that should be taken into consideration with both WBA and Boro's xG numbers is the fact that they have (probably) been behind in a lot of games this year. Game state can be a big factor in xG. As a match progresses, a team hanging on to a lead is going to drop deeper, probably give up more shots while having less themselves. Contrast this to us currently, when we have been leading going into the last part of the game, it's more likely we are the ones under pressure. Saying that however, most top sides will probably have a 2+ goal cushion and take the sting out of the game through possession, Man City being the masters at it.

this is an excellent point. I guess those xG graphs that Hoggy posts are good at fleshing out the bare stats.

[hogesar 9,618]

1 hour ago, Canaries north said:

Yes you can use this to prove your point but take a look at West Bromwich to see how these statistics can also be misleading. If you have Pukki and serve him he might not exceed XG but will be closer than almost all strikers in the division. Some people mention shots taken etc and that is a stupid stat and XG tries to make a better judgement of that and does but it's not perfect. XG shows the chances you are creating but it's the quality of the striker that gets you points. 

xG is far from perfect. xG, xGA are as close as we can really get to statistically analysing performances as fans, though.

And the key thing about xG and any other stat is it takes fan emotion out of the equation, which is what my OP was all about!

38 minutes ago, Robert N. LiM said:

this continues to be a really interesting thread and shows how stats can provide a good framework for discussion of how your team is playing. Cheers, @hogesar

Thanks. The more reading i've done by those who collate or utilise stats has me more interested. It's also interesting looking at it from a non-biased view in that there's been prem games i've watched thinking a certain team is on top but statistically the opposite team was under no real danger. Sometimes when we think managers have seen another game to us, it could well be because they're looking at it completely differently! 🙂

 

[repman 634]

Another point re West Brom is that their keeper, David Button, statistically is pretty much the worst keeper in the league. 

https://theanalyst.com/eu/2022/08/english-championship-stats-2022-23/

The link there is a great tool to look at some basic stats on teams/players.

Button has a goals prevented (xG conceded - Goals conceded) so far of -5.8.

Krul is 3rd with 2.6.

The top keeper so far according to the metric is.... John Ruddy with 3.9.

That being said it might be a little too soon to make a judgement (could all still be a bad run of luck for Bruce/WBA)

Edited October 4, 2022 by repman