Promoted Teams first season performance

  • Please bear with us on the new site integration and fixing any known bugs over the coming days. If you can not log in please try resetting your password and check your spam box. If you have tried these steps and are still struggling email [email protected] with your username/registered email address
  • Log in now to remove adverts - no adverts at all to registered members!
C

CBeeeee

Active Member
May 5, 2024
111
27
28
54
I had a bit of time so I thought I would try and give some perspective on expectations for next season given the historical records of previous promoted teams since the PL was created back in 1992. The idea is to try and measure not just a single best guess but to try and measure likely bounds of points and GD and also whether this has changed over time and whether Hull are different to other clubs. I downloaded final league tables and wrote some R code to analyse -code and data available if anyone wants it. I'll deal with Part 1 first, before looking at Hull's specific performance.

Firstly, there is incredible variation. I have taken points per game and then converted to a 38 game season so that we are comparing like for like, but even then we have a range from Derby's infamous 10 point season in 07/08 to Newcastle and Forest equaling 69.7 points in 93/94 and 94/95. The historical mean 38.8 and median 38 is incredibly close to what is usually regarded as the safety line (38 points)


You must log in or register to see images

Notably the high points seem to occur early in PL history 4 of the top five PPG occurred in the first 10 years with only Leeds (20/21) featuring more recently with 59 points. The reverse is also true, of the five worst returns, 3 of the worst five have occurred in the last six years (Norwich 19/20 - 21 pts ; Sheff U 23/24 - 16 points and Southampton 24/25 - 12 points).

Looking at average PPG per promoted team over time and fitting a spline to catch trends shows effectively 3 periods of PL.

You must log in or register to see images


An additional adjustment period about 92/93 to 04/05 which saw a systematic decrease in points achieved by the promoted teams; a period of stability 05/06 to 16/17 which saw a fairly consistent return of around 1 point per game for the promoted teams and then a third period 17/18 onwards in which it looks like the PL has become way more brutal to newly promoted teams - even with the fantastic returns Sunderland and Leeds had this season. Its an almost identical story for goal difference.

You must log in or register to see images


So the bottom line is that although newly promoted teams have a very wide distribution of points in the first season back, the PL is a far more brutal experience for newly promoted teams than it was 9 years ago when we were last promoted.
 
  • Like
Reactions: x, Cortez91, ashbees_captains_armband and 6 others
CBeeeee said:
I had a bit of time so I thought I would try and give some perspective on expectations for next season given the historical records of previous promoted teams since the PL was created back in 1992. The idea is to try and measure not just a single best guess but to try and measure likely bounds of points and GD and also whether this has changed over time and whether Hull are different to other clubs. I downloaded final league tables and wrote some R code to analyse -code and data available if anyone wants it. I'll deal with Part 1 first, before looking at Hull's specific performance.

Firstly, there is incredible variation. I have taken points per game and then converted to a 38 game season so that we are comparing like for like, but even then we have a range from Derby's infamous 10 point season in 07/08 to Newcastle and Forest equaling 69.7 points in 93/94 and 94/95. The historical mean 38.8 and median 38 is incredibly close to what is usually regarded as the safety line (38 points)


You must log in or register to see images

Notably the high points seem to occur early in PL history 4 of the top five PPG occurred in the first 10 years with only Leeds (20/21) featuring more recently with 59 points. The reverse is also true, of the five worst returns, 3 of the worst five have occurred in the last six years (Norwich 19/20 - 21 pts ; Sheff U 23/24 - 16 points and Southampton 24/25 - 12 points).

Looking at average PPG per promoted team over time and fitting a spline to catch trends shows effectively 3 periods of PL.

You must log in or register to see images


An additional adjustment period about 92/93 to 04/05 which saw a systematic decrease in points achieved by the promoted teams; a period of stability 05/06 to 16/17 which saw a fairly consistent return of around 1 point per game for the promoted teams and then a third period 17/18 onwards in which it looks like the PL has become way more brutal to newly promoted teams - even with the fantastic returns Sunderland and Leeds had this season. Its an almost identical story for goal difference.

You must log in or register to see images


So the bottom line is that although newly promoted teams have a very wide distribution of points in the first season back, the PL is a far more brutal experience for newly promoted teams than it was 9 years ago when we were last promoted.
Click to expand...
Good work. It confirms what a lot of people already think about the gap between the Premier League and the Championship getting wider in the last few years.
 
  • Like
Reactions: Idi Amin, balkan tiger and CBeeeee
Kalman II said:
Good work. It confirms what a lot of people already think about the gap between the Premier League and the Championship getting wider in the last few years.
Click to expand...
Its certainly consistent with that - its two things really that have driven the graph. The terrible returns in 3 of the last 5 years and the lack of really good returns recently. This year's average feels really high, but thats only because recently its been dismal. The 25/26 return is pretty much just historical average overall.
 
  • Like
Reactions: ashbees_captains_armband and Kalman II
If the gap between Championship and Prem is getting bigger does that mean if you do stay up each season it gets easier to stay up and if you do go down you can just go straight back up
 
TigerMarv said:
If the gap between Championship and Prem is getting bigger does that mean if you do stay up each season it gets easier to stay up and if you do go down you can just go straight back up
Click to expand...
From other research I did when I was bored, which was a few years ago and before the current increasing gap, yes (to part of your question). I estimated that the probability of relegation for a newly promoted team was very high (esp if via the play-offs) in the first season. And there was still an above average probability of relegation in the 2nd season but no-where near as high as in the first (so no second season syndrome then). If you survived 2 seasons then your relegation risk returned to what would be expected ie 3/20. So if you survive 2 seasons you become established. I haven't attempted the yo-yo question and that finding was drawn before the recent down-turn.

I did look at data since 1945 and it was sort of true pre-PL except you only needed one season survival to become established and the risk of first season relegation in old first division was almost exactly equal to the conditional second season survival in the PL. So the PL was already making it trickier for promoted teams to survive and thrive.
 
  • Like
Reactions: DarbyDays and TigerMarv
So how good are Hull, relative to other promoted teams? We have 3 previous promoted first seasons - 08/09, 13/14 and 16/17 which yielded 35, 37 and 34 points respectively. Slightly below historical means and very very little variation but only 3 observations which is not really a lot to go on. Indeed of the 41 clubs promoted the max number of observations is 5 first seasons which a few teams have (Sheff Utd, Sunderland, Leicester, WBA, Norwich, Burnley) but some just have 1 (Brentford, Wigan, Stoke, Blackpool and many more. So I'm going to use a common stats method which when allocating unit-specific effects gives us a fairly conservative estimate which is a weighted product of the overall mean and team-specific data. The more observations a team has the more the effect is determined by its specific history whereas a team which has just one obs will be more dragged towards the mean. Given that the max obs is 5 and that we can see a lot of variation the consequence is that this model will 'shrink' estimates towards the mean. This is generally seen as a positive property in models which are used for prediction. So I run such a model which tries to explain points as a function of the season (using the spline) and team-specific effect and the points per game achieved within the previous promotion campaign (so I can account for the quality of the team coming up).

To show raw team effects I have evaluated the expected team points in a mid-range season (2010) assuming the team was promoted the previous season with an average amount of points. The point of this is to get a good idea of the range of differences between teams.
expected team points in a mid-range season (2010) assuming the team was promoted the previous season with an average amount of points.

You must log in or register to see images


So the data and model predict that Hull are a below average team ranked 30/41. The top teams and very bottom teams feel expected. However note that I think this is quite conservative and a result of the 'shrinkage' I mentioned before whereby the model draws all teams closer to an average, more so the less data on individual teams. So for example Swindon are ranked 34 but had an appalling 27.1 pts (based on a 38 game season) in 93/94 when promoted teams where getting lots of points. So there is a good case for making Swindon the worst team, but because we only have 1 observation and there is a lot of variation it is really dragged/shrank towards an overall mean. And whilst this less true for teams with 5 obs (and thats why they dominate the bottom 6) I think the model really underestimates the expected differences between teams ie just a 5 point gap in expectation between the very worst, Burnley and the best, Leeds.
 
  • Like
Reactions: DarbyDays
The model predicts Hull's actual obsverations quite well. It has expectations of 37.5, 38.6 and 37.4 for the seasons in which we got 35, 37 and 34. So the model has slightly over-estimated our actual points obtained (which is consistent with me think that the model is a bit too conservative in thinking we are an average team.)
 
  • Like
Reactions: DarbyDays
but much less accurate for Sunderland say (though they are an absolute Jekyll and Hyde team) for example in their 05/06 season 40.6 points were estimated and only 15 gained. But anyway, here is the model prediction for points and GD for Hull's 26/27 campaign - based on Hull specific factor, relatively low points last season and most importantly of all, the brutal nature of the PL post 2017.


You must log in or register to see images
You must log in or register to see images


So the model has a point estimated of 25 points and - 51 GD. I think that seems broadly realistic and is almost entirely driven by how difficult the PL is these days. Hull being below average and with a low promoted points total has very little effect to be honest. The other thing to note is the huge spread - there is a 95% credible interval of 7 - 66 points absolutely vast. If promotion requires >= 38 points then there is a 15% chance we'll do that. And for >= 35 pts and >- 30 pts its 19% and 31% respectively. Current betting odds imply a 22% chance of survival so its consistent with the model.

So the bottom line from using historical data and a bit of mathematical modelling is that it is very unpredictable. The likelihood is that this is going to be much more difficult than last time as the PL has changed. But very little is certain and there is a reasonable chance of surviving and even a chance that we could thrive. The flip side of the uncertainty is that record low points are also on the table - the model predicting an 8% chance of doing worse than Derby's points and a 31% chance of having the worst GD on record.
 
  • Like
Reactions: ashbees_captains_armband
Few promoted clubs survive the gulf in the Premier League very long with many coming straight back down. No big deal and nothing new there. The good thing is that you usually return financially better off and of course with parachute payments. On a side note..People talk about how well Sunderland have done this season which is fair play but for me it’s more about how did the likes of Bournemouth,Brighton and Brentford have managed there affairs which quit frankly I find both refreshing and remarkable. Just goes to show you don’t have to be FC massive to survive the Gulf so It will be interested to see how City go about our business and potentially adjust or adapt facing the kind of talent you don’t see in the Championship but again..without knowing who we are going to sign that’s impossible to gauge at the moment.

UTT.
 
  • Like
Reactions: x and TigerMarv
Tentotwo said:
Few promoted clubs survive the gulf in the Premier League very long with many coming straight back down. No big deal and nothing new there. The good thing is that you usually return financially better off and of course with parachute payments. On a side note..People talk about how well Sunderland have done this season which is fair play but for me it’s more about how did the likes of Bournemouth,Brighton and Brentford have managed there affairs which quit frankly I find both refreshing and remarkable. Just goes to show you don’t have to be FC massive to survive the Gulf so It will be interested to see how City go about our business and potentially adjust or adapt facing the kind of talent you don’t see in the Championship but again..without knowing who we are going to sign that’s impossible to gauge at the moment.

UTT.
Click to expand...

I suppose just as you can show 3 teams who have gone up and managed to stay up for a few years there must be some teams that were up there and now aren't.
Are they in the financial poo because they came down after buying loads of more expensive players and stadium expansion?
 
balkan tiger said:
I suppose just as you can show 3 teams who have gone up and managed to stay up for a few years there must be some teams that were up there and now aren't.
Are they in the financial poo because they came down after buying loads of more expensive players and stadium expansion?
Click to expand...


Dunno mate.

Remember the likes of Spuds,Manure and Villa ect playing at Fer ark but they all went back. West Ham never got over moving from their old ground methinks. Burnley are a yo yo club but it’s not just about staying up though is it. The teams in question regularly playing excellent football and turn good sides over to boot. In Bo’muffs case they are doing it on 11K average gates apparently. Some promoted clubs never seem to get over having a crack at the Prem whilst one side actually won the Prem and have now been relegated yet again.

Bournemouth for me have thrown the rule book out the window.


UTT.
 
Tentotwo said:
Dunno mate.

Remember the likes of Spuds,Manure and Villa ect playing at Fer ark but they all went back. West Ham never got over moving from their old ground methinks. Burnley are a yo yo club but it’s not just about staying up though is it. The teams in question regularly playing excellent football and turn good sides over to boot. In Bo’muffs case they are doing it on 11K average gates apparently. Some promoted clubs never seem to get over having a crack at the Prem whilst one side actually won the Prem and have now been relegated yet again.

Bournemouth for me have thrown the rule book out the window.


UTT.
Click to expand...

Swansea, Cardiff, Derby, among others have been up there and now have fancy new stadiums.
Agree with Bournemouth they are showing it can be done without a fancy stadium and 10s of thousands of supporters. I would think that if they did get relegated they would be able to manage it well.
 
balkan tiger said:
Swansea, Cardiff, Derby, among others have been up there and now have fancy new stadiums.
Agree with Bournemouth they are showing it can be done without a fancy stadium and 10s of thousands of supporters. I would think that if they did get relegated they would be able to manage it well.
Click to expand...


Yup.

To think we were almost relegated last season and then get promoted the following year with several players involved at Pompey lifting the Play off winners trophy.

Best Game in the World mate.


UTT.
 
  • Like
Reactions: Kalman II, balkan tiger and TwoWrights
Tentotwo said:
Few promoted clubs survive the gulf in the Premier League very long with many coming straight back down. No big deal and nothing new there. The good thing is that you usually return financially better off and of course with parachute payments. On a side note..People talk about how well Sunderland have done this season which is fair play but for me it’s more about how did the likes of Bournemouth,Brighton and Brentford have managed there affairs which quit frankly I find both refreshing and remarkable. Just goes to show you don’t have to be FC massive to survive the Gulf so It will be interested to see how City go about our business and potentially adjust or adapt facing the kind of talent you don’t see in the Championship but again..without knowing who we are going to sign that’s impossible to gauge at the moment.

UTT.
Click to expand...
I think all the clubs you mentioned there had PL plans in place well before promotion - seasons before. I dont think A is that forward thinking.
 
You must log in or register to reply here.
Top Bottom
Back