Jump to content
IGNORED

Squad Churn


ExiledAjax

Recommended Posts

The transfer window slammed shut, and so we can say (with reasonable confidence bar any weird late free agent shenanigans etc) that Championship squads are now settled for the next few months. Experimental 361 is best know for his xG stuff, but he's really a data visualiser, and I think one interesting graphic he releases each season is the one below showing squad churn.

It shows that we have undergone about as drastic an overhaul as most of us would think. We've retained our three most played (in league games only) players from last season (in order of minutes played - Bentley (3,870), Kalas (3,383) and Vyner (3,364)) and have lost four of the top XI (Hunt, Rowe, Nagy and Diedhiou). Those four that we've lost played for an aggregate of 9,751 minutes last season and an average of 2,437*, that is a lot of minutes to replace. The other four from that most used XI are Wells, Semenyo, Bakinson and Martin. Note that this is the XI players that played the most minutes, it's not necessarily the players with the most starts, or those with the most starts together as an XI, but those were the XI guys that contributed most heavily to last season's league campaign.

Really though the big thing is that we shed a significant amount of the 'tail' of the squad. Outside that top XI we've retained only six of the others that played a significant number of league minutes. Those six are, from what I can see, Massengo, Palmer, COD, Dasilva, Weimann and O'Leary. We lost Mariappa, Moore, Paterson, Sessegnon, Mawson, Lansbury and Brunt. There's then quite a number of players who got fewer than 200 minutes over the season (people like Janneh, Towler, Williams etc) and they don't really show up well in the graphic.

*a 46 game league season is equal to 4,140 minutes.

2021-09-01-101.png?w=1000&h=

Comparing us to the rest of the division we see that we are in the top 6 "most-changed" squads. However, I think it is also pretty clear that, at least after 5 games, there's little correlation between how much a squad changes, and where you are in the table. Fulham are the most changed squad but are comfortably top - a symptom of relegation and losing loanees. Others such as Sheff Utd and Peterborough are much more stable, but are down in the bottom 6. There's really very little correlation that I can see.

Obviously the next step, and what isn't shown here, is to discuss who has been brought in to replace those that left. We basically got rid of 11 members of the first team squad, and brought in four. We also re-signed two of our "tail" of players (who bar injury would have probably been in the most used XI), but you can really see that we have stream-lined the squad. It also shows that we will need youth players like Scott, Janneh, Pring etc to come in and make up some of the minutes that we've lost. Now so far they are doing that pretty well, and long may it continue, but there is some pressure there to keep them doing well.

I guess really I don't know what this post says other than confirming that as we all know, we've streamlined the squad and promoted some youth in place of bringing in new players. Just nice to see it represented and compared to the rest of the division I guess.

Link to comment
Share on other sites

For me, always expected a fair bit of ‘churn’ this off-season, with the goal being having a squad where we can expect to see fewer players but more often.

The ‘Johnson Tombola’ was always in effect; sometimes it worked for us, but I think long term it worked against us, in that it involved needing a large number of similar players to facilitate that kind of approach.

If NP is a man of his word, this isn’t something he subscribes to, and his signings and emphasis on blooding and keeping youth involved seems to prove that, at least so far.

The hope will be that the ‘churn’ settles now for a year or two, with a younger core of the team being augmented by either more academy breakthroughs or the odd key signing, and those leaving being due to selling them on or letting them leave before they are past it, rather than holding on to loads of players.

Link to comment
Share on other sites

Interesting OP.

My view for what it’s worth is that with a few exceptions (Hunt, Rowe, Famara, plus arguably Paterson) we really haven’t shed too much by way of regular starters.

What we are clearly operating with is a far smaller squad with however in Nige’s view, still greater competition for places, but that is much more slimlined, so in central midfield for instance, James appears to be first choice & King & HNM are competing to partner him, but in Williams’ absence not half a dozen other potential possibilities.

As a consequence I think our starting line up will become increasingly easy to predict & Pearson’s mantra of team ethic & work rate clear.

This is a far more logical, cost effective way to operate than before.

Link to comment
Share on other sites

28 minutes ago, samo II said:

The hope will be that the ‘churn’ settles now for a year or two, with a younger core of the team being augmented by either more academy breakthroughs or the odd key signing, and those leaving being due to selling them on or letting them leave before they are past it, rather than holding on to loads of players.

I agree that hopefully both the management team and the matchday 18 largely settles down. Naturally a lot of that does ultimately depend on results and performances, but the hope is there.

I think the finances will dictate a little churn though. That is being discussed at length on other threads but it is pretty clear that (unless we get promoted) we are going to need some transfer profit of a fairly large magnitude either this season or next if we are to keep within FFP come 2023/24. That will create some churn, as of course will retirements, injuries etc.

Just to add - this is what we did at an equivalent time last season:

image.png.9d14fd2934a77a60de96e00b84f567fe.png

And at the start of 2019/20:

image.png.9cca67e91c8ad3238b9c4b22698883aa.png

And the start of 2018/19:

image.png.f39b527a8513f9558d67ab2ef01c0711.png

Those three show that whilst we've seen the biggest churn this summer, it's not too out of the ordinary when compared to previous summer windows. It's only slightly more, in terms of minutes 'lost', than we saw two years ago. I've not done a similar deep dive into exactly who left and who joined in each of these seasons, but you can see we tend to lose 2 or 3 of our most played players, and then a varying degree of the "tail" of the squad.

Link to comment
Share on other sites

The problem with stats like these us that they're heavily skewed by the injury situation at a club, and ours was worst than most. Undoubtedly Baker, Dasilva and most especially Weimann would have played many more minutes than they did if it wasn't for their long term injuries. Weimann would certainly have been in the top 5 of performance minutes. 

Link to comment
Share on other sites

59 minutes ago, ExiledAjax said:

I agree that hopefully both the management team and the matchday 18 largely settles down. Naturally a lot of that does ultimately depend on results and performances, but the hope is there.

I think the finances will dictate a little churn though. That is being discussed at length on other threads but it is pretty clear that (unless we get promoted) we are going to need some transfer profit of a fairly large magnitude either this season or next if we are to keep within FFP come 2023/24. That will create some churn, as of course will retirements, injuries etc.

Just to add - this is what we did at an equivalent time last season:

image.png.9d14fd2934a77a60de96e00b84f567fe.png

And at the start of 2019/20:

image.png.9cca67e91c8ad3238b9c4b22698883aa.png

And the start of 2018/19:

image.png.f39b527a8513f9558d67ab2ef01c0711.png

Those three show that whilst we've seen the biggest churn this summer, it's not too out of the ordinary when compared to previous summer windows. It's only slightly more, in terms of minutes 'lost', than we saw two years ago. I've not done a similar deep dive into exactly who left and who joined in each of these seasons, but you can see we tend to lose 2 or 3 of our most played players, and then a varying degree of the "tail" of the squad.

Be interesting to see how that compares on average to other sides across the last 3 seasons. Based on this seasons percentages we’d consistently be in the top quarter for player change.

Does the amount of ‘churn’ we see every season not indicate that for so long we’ve not known what we’re doing recruitment wise? Wholesale changes every season and not adding the odd bit of quality here and there?

LJ said himself a month or so back that he wanted 6 new players at Sunderland, OR 3 quality players. Mark Ashton has just overseen 18 new players to Ipswich. If ever you needed confirmation that we didn’t ever have a set system to tweak and improve.

Im more than happy that we’ll only recruit if it’s better than what we have, and it will be interesting to see this churn percentage in the next few seasons under NP as he forms his team

Link to comment
Share on other sites

14 minutes ago, RoystonFoote'snephew said:

The problem with stats like these us that they're heavily skewed by the injury situation at a club, and ours was worst than most. Undoubtedly Baker, Dasilva and most especially Weimann would have played many more minutes than they did if it wasn't for their long term injuries. Weimann would certainly have been in the top 5 of performance minutes. 

This will be a factor yes, and I noted that Baker and Weimann would almost certainly have been up there for minutes played (Weimann has played 450 and Baker 284 of the maximum 450 league minutes so far this season). Dasilva has 369 so far this term. However, honestly it's beyond my time and energy to weight all of this stuff to account for injuries. I'm not really sure how much it would really add either. We Personally I am happy to review these figures as they are, and just keep in mind that some players who might have gather 3,000+ minutes didn't due to injury.

The bit in bold - honestly I don't know enough details of the other 23 teams' injury records to really say that for certain we had it so much worse than them that it will heavily skew our squad churn statistics when compared to the rest of the division. I know anecdotally last season was bad, but most of us on here know about Bristol City's injuries, major or notable injuries at other clubs, but not much more than that.

13 minutes ago, deadredfred said:

Does the amount of ‘churn’ we see every season not indicate that for so long we’ve not known what we’re doing recruitment wise? Wholesale changes every season and not adding the odd bit of quality here and there?

It's hard to truly compare each season, as each season there are different factors at play. For example this summer we saw a lot of players run their contract out, whereas in 2018 we had three very saleable assets in Bryan, Reid and Flint. So there are different reasons why we 'lost' a lot of minutes in each of those summer windows. I'm therefore reticent to really use this data to analyse our overall recruitment strategy - personally I think that is better done through financial analysis, and the analysis of each individual player's performance whilst at the club. I look at this churn data at pretty much face value.

Link to comment
Share on other sites

6 hours ago, ExiledAjax said:

The transfer window slammed shut, and so we can say (with reasonable confidence bar any weird late free agent shenanigans etc) that Championship squads are now settled for the next few months. Experimental 361 is best know for his xG stuff, but he's really a data visualiser, and I think one interesting graphic he releases each season is the one below showing squad churn.

It shows that we have undergone about as drastic an overhaul as most of us would think. We've retained our three most played (in league games only) players from last season (in order of minutes played - Bentley (3,870), Kalas (3,383) and Vyner (3,364)) and have lost four of the top XI (Hunt, Rowe, Nagy and Diedhiou). Those four that we've lost played for an aggregate of 9,751 minutes last season and an average of 2,437*, that is a lot of minutes to replace. The other four from that most used XI are Wells, Semenyo, Bakinson and Martin. Note that this is the XI players that played the most minutes, it's not necessarily the players with the most starts, or those with the most starts together as an XI, but those were the XI guys that contributed most heavily to last season's league campaign.

Really though the big thing is that we shed a significant amount of the 'tail' of the squad. Outside that top XI we've retained only six of the others that played a significant number of league minutes. Those six are, from what I can see, Massengo, Palmer, COD, Dasilva, Weimann and O'Leary. We lost Mariappa, Moore, Paterson, Sessegnon, Mawson, Lansbury and Brunt. There's then quite a number of players who got fewer than 200 minutes over the season (people like Janneh, Towler, Williams etc) and they don't really show up well in the graphic.

*a 46 game league season is equal to 4,140 minutes.

2021-09-01-101.png?w=1000&h=

Comparing us to the rest of the division we see that we are in the top 6 "most-changed" squads. However, I think it is also pretty clear that, at least after 5 games, there's little correlation between how much a squad changes, and where you are in the table. Fulham are the most changed squad but are comfortably top - a symptom of relegation and losing loanees. Others such as Sheff Utd and Peterborough are much more stable, but are down in the bottom 6. There's really very little correlation that I can see.

Obviously the next step, and what isn't shown here, is to discuss who has been brought in to replace those that left. We basically got rid of 11 members of the first team squad, and brought in four. We also re-signed two of our "tail" of players (who bar injury would have probably been in the most used XI), but you can really see that we have stream-lined the squad. It also shows that we will need youth players like Scott, Janneh, Pring etc to come in and make up some of the minutes that we've lost. Now so far they are doing that pretty well, and long may it continue, but there is some pressure there to keep them doing well.

I guess really I don't know what this post says other than confirming that as we all know, we've streamlined the squad and promoted some youth in place of bringing in new players. Just nice to see it represented and compared to the rest of the division I guess.

That's a nice bit of work Ajax. Thanks.

Link to comment
Share on other sites

@ExiledAjax I think the squad churn metric is most useful purely looking at sides without parachute payments who were championship teams last season.

Admittedly that's a not insignificant number of teams excluded, but I suspect that come the end of the season there will be a soft correlation out of those teams that the ones with least churn will generally do better, though there will be outliers.

Link to comment
Share on other sites

6 hours ago, JamesBCFC said:

@ExiledAjax I think the squad churn metric is most useful purely looking at sides without parachute payments who were championship teams last season.

Admittedly that's a not insignificant number of teams excluded, but I suspect that come the end of the season there will be a soft correlation out of those teams that the ones with least churn will generally do better, though there will be outliers.

You could focus on the middle 18, removing the 3 Lg1 promoted teams and the 3 relegated PL ones couldn’t you?

Link to comment
Share on other sites

6 hours ago, JamesBCFC said:

@ExiledAjax I think the squad churn metric is most useful purely looking at sides without parachute payments who were championship teams last season.

Admittedly that's a not insignificant number of teams excluded, but I suspect that come the end of the season there will be a soft correlation out of those teams that the ones with least churn will generally do better, though there will be outliers.

Agreed. Promoted and relegated teams have unique circumstances that make it less relevant.

35 minutes ago, Davefevs said:

You could focus on the middle 18, removing the 3 Lg1 promoted teams and the 3 relegated PL ones couldn’t you?

The guys done this for the past 4 or 5 seasons so we can see if there's any correlation with the final tables there. 

Link to comment
Share on other sites

14 hours ago, JamesBCFC said:

@ExiledAjax I think the squad churn metric is most useful purely looking at sides without parachute payments who were championship teams last season.

Admittedly that's a not insignificant number of teams excluded, but I suspect that come the end of the season there will be a soft correlation out of those teams that the ones with least churn will generally do better, though there will be outliers.

Ok, some lunch break stats here. I looked at the last 4 seasons of the Championship and listed out the final finishing positions, together with the squad churn (expressed as a % of the previous season's minutes retained) numbers taken from E361. We get the below rough table.

I have highlighted in each season the teams that were relegated from the Prem or promoted from L1. I've then got four averages at the bottom of each column. The first is the overall divisional average, then the average for the 3 promoted teams, then the three relegated Premier League teams, and then the 18 teams that competed in the Championship in the prior season.

I'm saying right here - in general, in the Championship, squad churn % on 1 September is no indication of likely final league place.

However, I would caveat that by saying that generally speaking a high degree of churn should not, in itself, worry a fan. Likewise, a very low level should not worry a fan. However, in general teams that do well also lose somewhere around 30-35% of their minutes from the prior season.  

image.thumb.png.ae3cdb43d5ae487e2d69671f75848960.png

Honestly, of most interest to me are the averages. Especially comparing the Covid hit 2020/21 season to the preceding 3. We see a big jump in the overall level of retention, and that is across the board - promoted, relegated, and the rest all retained nearly 10% more minutes in that window than they did previously. That's a real representation of the depressed market that we saw last summer.

This season's overall average is 68.5 btw, so it's regressed a little back towards normal levels. Sheff Utd, WBA and Fulham have an average of 62.3 and Posh, Blackpool and Hull are up at 76.5. So again those two figures are more like the seasons prior to Covid rather than those from last year's truncated and financially unpredictable summer window.

One thing I might just say here, and I do not know the answer is, to ask someone like @Davefevs whether, after seeing this depression in the transfer market last year, can we better understand why we had a policy last summer of, for example, retaining Diedhiou and others who were in the final year of contract in 2020/21? Do we see here that in our division there was a general mood of "hold on to what you've got" rather than undergo transfer, and all of the costs and uncertainty they bring?

I can't do more than this over my lunch break, but might look at the other EFL divisions if I get time over the weekend.

Link to comment
Share on other sites

You read my mind @ExiledAjax….I suspect the lower churn this summer from last summer is down to contracts expiring.  There seemed to be a load of players released, of which there was a dearth of real quality in the group of players I thought might be realistic….perhaps indicating in the main they weren't hogging minutes last season, and the players kept were those who played.  Does that make sense?

Link to comment
Share on other sites

31 minutes ago, Davefevs said:

You read my mind @ExiledAjax….I suspect the lower churn this summer from last summer is down to contracts expiring. No, we've seen a slightly higher churn this summer. More movement, not less, at least in terms of minutes played. Sorry, the way E361 expresses the data is slightly backwards. The % is the figure retained, so a lower % number indicates more churn. 100% would indicate not losing a single player. There seemed to be a load of players released, of which there was a dearth of real quality in the group of players I thought might be realistic….perhaps indicating in the main they weren't hogging minutes last season, and the players kept were those who played.  Does that make sense?

One thing that the graphic I first posted doesn't tell you - at least not directly - is the raw number of players that leave each club. A team that loses it's 5 most played players will have a very high churn % even though they have lost only 5 players. Likewise a team could lose 15 players, but if each played only 3 games then their churn might look smaller than the other club I describe. Which of those has the biggest impact on a dressing room -  I don't know, and maybe that's a failing of this type of analysis.

What we did see last summer was (generally speaking) teams holding on to their "core". In the summer between 2019/20 and 2020/21, bar two of the relegated Premier League sides, QPR, Forest, and Huddersfield, very few teams lost more than one player who was in their top 6 for minutes played in the 2019/20 season. 10 didn't lose a single one of their top 6. See below.

2020-10-17-2-ch.png?w=860&h=792

But, if you look at my OP then you'll see that's a bit different this season. Just looking at it, I'd say that at least half the teams have lost 2 or more of their top 6, and I think there's inly 5 that lost none.

So I'd say that on the face of it this backs up what I think you're saying - that in summer 2020 we saw a lot of clubs extend contracts, or retain players that they might previously have sold, rather than cash in on a key player. Diedhiou is probably a text book example this. 

Link to comment
Share on other sites

11 hours ago, ExiledAjax said:

Ok, some lunch break stats here. I looked at the last 4 seasons of the Championship and listed out the final finishing positions, together with the squad churn (expressed as a % of the previous season's minutes retained) numbers taken from E361. We get the below rough table.

I have highlighted in each season the teams that were relegated from the Prem or promoted from L1. I've then got four averages at the bottom of each column. The first is the overall divisional average, then the average for the 3 promoted teams, then the three relegated Premier League teams, and then the 18 teams that competed in the Championship in the prior season.

I'm saying right here - in general, in the Championship, squad churn % on 1 September is no indication of likely final league place.

However, I would caveat that by saying that generally speaking a high degree of churn should not, in itself, worry a fan. Likewise, a very low level should not worry a fan. However, in general teams that do well also lose somewhere around 30-35% of their minutes from the prior season.  

image.thumb.png.ae3cdb43d5ae487e2d69671f75848960.png

Honestly, of most interest to me are the averages. Especially comparing the Covid hit 2020/21 season to the preceding 3. We see a big jump in the overall level of retention, and that is across the board - promoted, relegated, and the rest all retained nearly 10% more minutes in that window than they did previously. That's a real representation of the depressed market that we saw last summer.

This season's overall average is 68.5 btw, so it's regressed a little back towards normal levels. Sheff Utd, WBA and Fulham have an average of 62.3 and Posh, Blackpool and Hull are up at 76.5. So again those two figures are more like the seasons prior to Covid rather than those from last year's truncated and financially unpredictable summer window.

One thing I might just say here, and I do not know the answer is, to ask someone like @Davefevs whether, after seeing this depression in the transfer market last year, can we better understand why we had a policy last summer of, for example, retaining Diedhiou and others who were in the final year of contract in 2020/21? Do we see here that in our division there was a general mood of "hold on to what you've got" rather than undergo transfer, and all of the costs and uncertainty they bring?

I can't do more than this over my lunch break, but might look at the other EFL divisions if I get time over the weekend.

So for your %s is a higher number more minutes from the previous season retained?

Link to comment
Share on other sites

9 hours ago, Davefevs said:

Yes

Cheers, was going to do scatterplots for the 18 teams who were Championship the season before from each of the tables, then a final one using average "churn" and average finishing positions.

Mostly because it helps me see what's being shown more than a table alone

Link to comment
Share on other sites

2 hours ago, JamesBCFC said:

Cheers, was going to do scatterplots for the 18 teams who were Championship the season before from each of the tables, then a final one using average "churn" and average finishing positions.

Mostly because it helps me see what's being shown more than a table alone

I'd be interested to see that. If you want access to the spreadsheet I've prepped then DM me. Might save you 30 minutes of data entry.

Link to comment
Share on other sites

1 hour ago, JonDolman said:

The new players have made a difference, but the biggest difference is having players fit.

I think some might have been thinking we need to bring in loads of new players. But the end of last season a fit squad could have meant Baker and Kalas as a centre back pairing. Williams and Massengo as a midfield pairing. Dasilva would have been at left back. Pring had been recalled so Nige could have had him on the wing like he is now, Semenyo and O'Dowda options too. And Martin and Weimann would have been available up top.

So a hugely improved side would have been playing. And actually we could add Mawson to that list as he'd still have been there if he had not got injured. And Pato fit would have improved us too.

The players that were let go I think Pearson got spot on. And the signings have made us better for sure. But we didn't need much in at all. Maybe missing that striker to compete with Martin as we play a lot of high balls up to Martin. Will be interesting to see how we play if Wells comes in for him some games.

The new right back could be a clever bit of business. Vyner had a very solid game against Cardiff and made some very good forward passes. Simpson been fine in the league so far. Not too much of a problem there but great to have a more attacking option who will no doubt be taking over from Simpson in the future.

….and yet people were saying we needed 10-12 new signings when the retained / released list was announced.  Was never gonna happen.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...