Does DaveFevs Have The Solution?

[Galley is our king]

4 hours ago, BTRFTG said:

Despite all the positives last night we were again ultimately undone by woeful distribution in the final third. Dead ball or in play City's inability to clear the first man, pick out a team mate, avoid passing to space, avoid over-hitting the pass or even managing to keep the thing in play - what stats are there to show where City compare with others in the EFL? I find it difficult to identify who City's worst culprit in this respect as they all appear equally inept at playing a final, quality pass. Puzzling given they all play decent football in getting to that point. Probably also something to do with positioning and movement of strikers in the box but its incredible how few decent chances we create compared to those possession and positioning suggest we should.

Maybe this answers some supporters question as to why Kal Naismith, a centre half,  takes corners and free kicks?

[cheese]

25 minutes ago, maxjak said:

Sheffield Utd didn't make their own luck...........they had a referee to do it for them?...........When will we have a Ref to do that for us?  What with the highest number of games without a penalty, added   to the men in black missing offsides and red card offences, it is not so much bad luck, as a curse.  It all balances out over a season i hear people say...............well i have not seen much pheckin balance,  ever since Hull stole a penalty in the 1st game of our season!! 

 

The luck is being saved for 2030 when we get awarded a dodgy penalty in the 90th minute of the Champions League Final to win 1-0.

[RedRock]

Was just having a chat with my hairdresser, who was a good standard of footballer in his time.

Both of us agreed, a cross is far more effective from the bye line, cutting it back away from the goal to advancing forwards. Modern day football with wing-backs, has a tendency for crosses to come in from the edge of the penalty box - at best. Most crosses come in from a zone back from that point towards the half way line. City seem to be particularly adept to these long range ‘passive’ crosses.

The one last night in the second half - I think Sykes - which caused havoc was from near the bye line with pace. Wells should have been on the end of that. 

May be a time for a reset in our forward play. 

[BTRFTG]

22 minutes ago, SBB said:

New to me but behind the ball doesn’t matter anymore and is no longer in the rules. Tottenham’s disallowed goal in the champions league last week taught me that. 

Wrong.

"A player is in an offside position if:

any part of the head, body or feet is in the opponents’ half (excluding the halfway line)

and

any part of the head, body or feet is nearer to the opponents’ goal line than BOTH the BALL and the SECOND-LAST OPPONENT."

If the ball is nearer the opponent's goal line a player behind it is NOT in an offside position.

 

[Major Isewater]

30 minutes ago, Mr Popodopolous said:

 

Tone set early.

Did he by Jove? 

[Mr Popodopolous]

1 minute ago, Major Isewater said:

Did he by Jove? 

Bar not open yet, Major?

(Sure that was a Fawlty Towers line).

Edited November 2, 2022 by Mr Popodopolous

[SBB]

32 minutes ago, 1960maaan said:

No, Kane was offside when Emerson (?) heads the ball. The stupid part of the rule is not , it doesn't matter if you're behind the ball. The stupid part is it dose,'t matter what direction the ball is going. 
And after decades of "you can't be offside if you play the ball backwards"

Yeah I got the situation confused with direction of travel. Cheers

22 minutes ago, BTRFTG said:

Wrong.

"A player is in an offside position if:

any part of the head, body or feet is in the opponents’ half (excluding the halfway line)

and

any part of the head, body or feet is nearer to the opponents’ goal line than BOTH the BALL and the SECOND-LAST OPPONENT."

If the ball is nearer the opponent's goal line a player behind it is NOT in an offside position.

 

As above. Thanks

[Jerseybean]

3 hours ago, BTRFTG said:

Thanks, not I'm sure I understand any of it. My point was somewhat rather more expansive in it seems to me that beyond most clubs we regularly get the ball into good areas of the opponent's half and then simply cede possession without creating threat. I guess somebody somewhere measures this but interested to hear what others think?

For example, how many dead balls are first touched by opposition? Are there crossing stats; how many under/over hit, hit behind, hit to opponents et al?

Regardless of any stats I simply look at our football and compare it with that of our opponents, the recurring, albeit non-evidenced themes (as in not substantiated by data/stats), are:

(1) we lose possession very often (compared with most opponents)

(2) when we get in and around the opponents box we are less likely to pass it around opting either to be direct or go backwards (there was a passage of play last night when SU played the ball across and around our box for what felt like ages awaiting an opening)

(3) our distribution from throw ins, free kicks and corners is generally poor and we frequently concede possession from promising positions

(4) we win very few 50/50 challenges 

(5) our forwards rarely win headers from long through balls whether from the keeper or defenders

All this adds up to great frustration! 

[grifty]

46 minutes ago, Mr Popodopolous said:

Bar not open yet, Major?

(Sure that was a Fawlty Towers line).

Papers arrived Fawlty?

No, not yet Major

[Red-Robbo]

If the solution is to start out-of-form Chris Martin and give him a run of games then I beg to differ.

He has one goal in 9 hours of football this season. I expect he may start (and may score) against Lincoln, but if we're talking about second tier, he's gone off the boil and the same rules must apply to him as they apply to Bentley and any other out-of-form player who have to wait until they get another chance to force their way back into contention. In his case, with goals.

[Davefevs]

3 hours ago, Mr Popodopolous said:

XG using the Experimental 361 had us winning.

1.8-0.5 in our favour.

Wyscout was BC 1.32 v SU 0.62

[ExiledAjax]

3 minutes ago, Davefevs said:

Wyscout was BC 1.32 v SU 0.62

1.62 v 0.8 from my average of 4 sources. 

[REDOXO]

2 hours ago, maxjak said:

Sheffield Utd didn't make their own luck...........they had a referee to do it for them?...........When will we have a Ref to do that for us?  What with the highest number of games without a penalty, added   to the men in black missing offsides and red card offences, it is not so much bad luck, as a curse.  It all balances out over a season i hear people say...............well i have not seen much pheckin balance,  ever since Hull stole a penalty in the 1st game of our season!! 

 

No one will disagree.
 

But crying about it won’t change it. The solution is to take our opportunities and stop giving away soft touch goals. Pearson alluded to it. Make your own luck and we may get a few more decisions! 

[Simon79]

3 hours ago, And Its Smith said:

I don’t think he shot. I think it was a misplaced cross. 

That was how I saw it when watching it on Sky, just couldn’t get his foot around it enough. COYR 

[Rocking Red Cyril]

Surely the only stat that matters is the goals

Bristol city 0 Sheffield utd. 1

[Sleepy1968]

1 hour ago, Jerseybean said:

(3) our distribution from throw ins, free kicks and corners is generally poor and we frequently concede possession from promising positions

I wish it was. It would be two or three steps up from what it is now.

[Mr Popodopolous]

40 minutes ago, Rocking Red Cyril said:

Surely the only stat that matters is the goals

Bristol city 0 Sheffield utd. 1

In many ways yes, although underlying performance can be an indicator of things to come, whether good or bad.

Edited November 2, 2022 by Mr Popodopolous

[Rocking Red Cyril]

3 hours ago, Mr Popodopolous said:

In many ways yes, although underlying performance can be an indicator of things to come, whether good or bad.

Yes I can see the point. But we better hope we soon transfer the good stats into goals and get some positive results and points or this is turning into a relegation season . And I hate that thought and the fact I am even saying it but .............

I feel there has to be movement in and out with players in Jan. Window

Again just my thoughts. 

[Lrrr]

5 hours ago, Davefevs said:

Wyscout was BC 1.32 v SU 0.62

Can’t say I overly trust wyscout’s xG model 

[firstdivision]

16 hours ago, Davefevs said:

If you’re really interested in stats, there are paid for services like Wyscout (relatively cheap for data, video not so), Statsbomb (extortionate), or free services like FBRef who’ve just swapped from getting a subset of data from Statsbomb to using Opta.

But with all data (football data) the art is understanding the context in which the data is collected, and how you can piece bits together to form a hypothesis.

One person I know of flits from using possession one week, to shots another, to shots on target another to frame their agenda.  Last night shows the problem with doing that.  We won the possession, we won the shots, we drew the shots on target, but two of our off target shots crashed against the woodwork.  You can’t just use stats in isolation, unless you really want to look at the scoreline!  That is of course the one that matters.

FBref:

https://fbref.com/en/comps/10/Championship-Stats

Heres the kind of data you can look at:

 

The more of Net Gains I read, Dave, the more it becomes apparent that everyone’s understanding of football stats is still far from complete. We’re still fumbling around in the dark to some extent because of the nature of the game - its fluidity and scoring system. 
It seems the best use might be for uncovering undervalued talent. 

[BTRFTG]

13 hours ago, Red-Robbo said:

 I expect he may start (and may score) against Lincoln,

"The magic of the cup..." (sic) Spare me . .

I neglected to spot the draw had been made for the 3rd round let alone forced onto a Tuesday night, early November.

And to think the FA Cup once mattered ....

 

[Davefevs]

7 hours ago, Lrrr said:

Can’t say I overly trust wyscout’s xG model 

It is what it is, it’s better than infogol, but not as good as Statsbomb.

1 hour ago, firstdivision said:

The more of Net Gains I read, Dave, the more it becomes apparent that everyone’s understanding of football stats is still far from complete. We’re still fumbling around in the dark to some extent because of the nature of the game - its fluidity and scoring system. 
It seems the best use might be for uncovering undervalued talent. 

There is also a difference between “understanding” and then “using” any stats.

If a successful pass is counted purely because the next contact is by our player, you can have a player who gets a tick on the box even though it’s a hospital pass.  Nagy was a good example of high pass success, but the pass receiver was sometimes played into trouble.  That’s difficult to quantify by stats.  In my head there’s a model that needs to show “what happens next” and then link that back, e.g. Nagy passes to 5 players with 90% success, but each of those 5 players have varying stats for retaining the ball.  Does that highlight Nagy giving hospital passes, or does it highlight the receiving player deficiency.  I’m a big believer in player pairings, combos, e.g. how does the RWB link with the CM and the RCB3.  I digress.

Football is far too fluid a game (as you state) to rely solely on stats, the eyes are still way more important.  I like stats to be the “backing up” of what the eyes saw.  And sometimes you’re surprised, although we shouldn’t be.  In the main, us fans are “ball watchers”, we aren’t looking at off the ball stuff, shape changes etc.  You can of course use stats for trends.

You are right, there is great opportunity to use stats to streamline the search for new players.

Take Joe Ward at Peterborough for example.  We are looking for a creative RWB.

Just knocking up a set of scatter diagrams shows that he does a lot of (x axis):

- passes to the pen area

- through passes

- crosses

(Ignore free-kicks / corners)

compared to all Lg1 FBs/WBs

But, how accurate (y-axis) is he:

- average success for passes

- slightly above average for through passes

- average for crosses

You can obviously repeat this for many other “stats”….I have an additional 17 (plus the 4 above) of these that help me build a data picture of a player, both with and without ball.

I’d still have lots of Qs, like:

- how much of his volume is because of the team he plays for and their style of play

So, can help you fine tune, but it can’t tell me everything, and nor do I ever start by looking at data…unless someone asks me - “who is the best ball winner in Lg1/2 who can pass well”.  You’d still caveat it massively!

 

[spudski]

One stat I'd love to find out, which I can't find, and I'm not even sure whether it's ever been looked at...is what formation is the most successful.

I know it's subjective...and sides play differently, regardless of how they set up...however there is a link.

I really don't like playing with 3 at the back and ' wingbacks'. Imo it's one of the weakest formations at this level and above. 

Where as 433 with full backs that drive forward to support, I feel is just as effective going forward, but more secure at the back.

 

[Davefevs]

Just now, spudski said:

One stat I'd love to find out, which I can't find, and I'm not even sure whether it's ever been looked at...is what formation is the most successful.

I know it's subjective...and sides play differently, regardless of how they set up...however there is a link.

I really don't like playing with 3 at the back and ' wingbacks'. Imo it's one of the weakest formations at this level and above. 

Where as 433 with full backs that drive forward to support, I feel is just as effective going forward, but more secure at the back.

 

It will change over time.  Certain formations become en vogue, trying to mirror certain teams.  We only have to go back to 66 and Alf’s 442 wingless wonders, and see how that changed how so many team played.  And even within that, you have different styles.

Re your 433 and how you’d play it, I see that as your “philosophy”, and if you were manager, you’d have a good idea of how you’d want to develop and recruit your squad to play that way.  The squad you start with might be a million miles from it though, and that’s where the fun begins!

[spudski]

5 minutes ago, Davefevs said:

It will change over time.  Certain formations become en vogue, trying to mirror certain teams.  We only have to go back to 66 and Alf’s 442 wingless wonders, and see how that changed how so many team played.  And even within that, you have different styles.

Re your 433 and how you’d play it, I see that as your “philosophy”, and if you were manager, you’d have a good idea of how you’d want to develop and recruit your squad to play that way.  The squad you start with might be a million miles from it though, and that’s where the fun begins!

Yes I realise it has changed over time Dave. And what becomes envogue etc.

However...I was hoping there would be some sort of evidence as to what worked best on a yearly basis. 

And yes... I can see that we most likely have tried to find the best way of playing with the players we have at our disposal.

Have recent NP additions been bought in to fit with our present set up? Arguably yes.

Are our academy players playing the same way? Yes.

Is player pathway being looked at, so as to fit in a seemless way? Yes.

Are potential future signings being looked at to fit in the same way...I guess so. ( Interesting we were after the young German lad that came off early for Sheff Utd. )

I guess my point being...is this playing style with wide wing backs and a reliance on crossing and pressing our ' Identity'...a footballing philosophy for the club throughout etc?

If so....I'm not sure it's a good one.

Or can we adapt. Invert the wing backs and create more narrow etc.

It's something I look at and wonder.

[Davefevs]

Just now, spudski said:

Yes I realise it has changed over time Dave. And what becomes envogue etc.

However...I was hoping there would be some sort of evidence as to what worked best on a yearly basis. 

And yes... I can see that we most likely have tried to find the best way of playing with the players we have at our disposal.

Have recent NP additions been bought in to fit with our present set up? Arguably yes.

Are our academy players playing the same way? Yes.

Is player pathway being looked at, so as to fit in a seemless way? Yes.

Are potential future signings being looked at to fit in the same way...I guess so. ( Interesting we were after the young German lad that came off early for Sheff Utd. )

I guess my point being...is this playing style with wide wing backs and a reliance on crossing and pressing our ' Identity'...a footballing philosophy for the club throughout etc?

If so....I'm not sure it's a good one.

Or can we adapt. Invert the wing backs and create more narrow etc.

It's something I look at and wonder.

Yeah, not seen any stats per se.

Re City, it’s clear from recent comments, Nige isn’t a back 3 man either, but he’s compromised because of the players he has. My big question is - how does he transition this squad to a back 4 system?  My second biggest question is - how does that reflect on previous recruitment?

My thoughts - why bring in wingbacks (Wilson)?  Why recruit a ball-playing CB (Naismith) when you want a ball winner?

My answer - he got what he could and adapted, and in an ideal world he’d have brought in different players.

For pure speculation, had we signed Rinomhota (or similar) as our first signing, ie bringing in a ratty CM, might he have felt better about going with 2 CBs and using Rinomhota as the screen?  The rest of recruitment might’ve revolved around whether he got Naismith or Rinomhota.  Getting both wasn’t necessary / affordable.

I’m intrigued where Khadra was gonna fit in….maybe it was as cover for someone leaving like Antoine?

[spudski]

7 minutes ago, Davefevs said:

Yeah, not seen any stats per se.

Re City, it’s clear from recent comments, Nige isn’t a back 3 man either, but he’s compromised because of the players he has. My big question is - how does he transition this squad to a back 4 system?  My second biggest question is - how does that reflect on previous recruitment?

My thoughts - why bring in wingbacks (Wilson)?  Why recruit a ball-playing CB (Naismith) when you want a ball winner?

My answer - he got what he could and adapted, and in an ideal world he’d have brought in different players.

For pure speculation, had we signed Rinomhota (or similar) as our first signing, ie bringing in a ratty CM, might he have felt better about going with 2 CBs and using Rinomhota as the screen?  The rest of recruitment might’ve revolved around whether he got Naismith or Rinomhota.  Getting both wasn’t necessary / affordable.

I’m intrigued where Khadra was gonna fit in….maybe it was as cover for someone leaving like Antoine?

Those are all questions I've asked myself as well.

Because like you point out...from the outside look odd. Especially Naismith and Wilson.

I hope it's more than make do and mend. 

Khadra is also an intriguing fit. Especially when he's known for poor work rate and defending. And awful at crossing.

[firstdivision]

4 hours ago, Davefevs said:

It is what it is, it’s better than infogol, but not as good as Statsbomb.

There is also a difference between “understanding” and then “using” any stats.

If a successful pass is counted purely because the next contact is by our player, you can have a player who gets a tick on the box even though it’s a hospital pass.  Nagy was a good example of high pass success, but the pass receiver was sometimes played into trouble.  That’s difficult to quantify by stats.  In my head there’s a model that needs to show “what happens next” and then link that back, e.g. Nagy passes to 5 players with 90% success, but each of those 5 players have varying stats for retaining the ball.  Does that highlight Nagy giving hospital passes, or does it highlight the receiving player deficiency.  I’m a big believer in player pairings, combos, e.g. how does the RWB link with the CM and the RCB3.  I digress.

Football is far too fluid a game (as you state) to rely solely on stats, the eyes are still way more important.  I like stats to be the “backing up” of what the eyes saw.  And sometimes you’re surprised, although we shouldn’t be.  In the main, us fans are “ball watchers”, we aren’t looking at off the ball stuff, shape changes etc.  You can of course use stats for trends.

You are right, there is great opportunity to use stats to streamline the search for new players.

Take Joe Ward at Peterborough for example.  We are looking for a creative RWB.

Just knocking up a set of scatter diagrams shows that he does a lot of (x axis):

- passes to the pen area

- through passes

- crosses

(Ignore free-kicks / corners)

compared to all Lg1 FBs/WBs

But, how accurate (y-axis) is he:

- average success for passes

- slightly above average for through passes

- average for crosses

You can obviously repeat this for many other “stats”….I have an additional 17 (plus the 4 above) of these that help me build a data picture of a player, both with and without ball.

I’d still have lots of Qs, like:

- how much of his volume is because of the team he plays for and their style of play

So, can help you fine tune, but it can’t tell me everything, and nor do I ever start by looking at data…unless someone asks me - “who is the best ball winner in Lg1/2 who can pass well”.  You’d still caveat it massively!

 

That reply was beyond the call of duty, Dave.

Yes, when I wrote 'understanding' I meant it in its widest sense - trying to work out what stats are helpful, trying to ask the right question in the first place, trying to work out what the data means, trying to work out what the learning point is in the data etc.

The things I've 'learnt'(had confirmed) are: there's a hell of a lot we don't know how to interpret (yet); wages are important; some players are undervalued - try to find them; a successful set piece strategy is important; no one quite knew why Busquets was such a good player but he was; and - this is me extrapolating - Antoine's misses against Swansea and Sheffield United were not all bad. At least he was there on the end of great chances. This plays into one of the ideas - that most decent strikers are quite similar in their ability to shoot where they want, but it's the quality of the chance, and being able to find the space to be there for it, that is most important.