Tranmere Proposal for Season End | Vital Football

Tranmere Proposal for Season End

I found it very interesting reading, and good statistically. However, their resolution to exclude teams from relegation if proved statistically unsound by their system would result in an imbalance in divisional numbers, specifically 25 teams in League one and 23 in League two.

So I would add an addition to that proposal, that teams who should not be relegated in the analysis should be required to prove their worthy status of keeping their position by taking part in the promotion play-offs from the division below.

When the current Play-offs first began, from memory (please correct me if I am wrong) I believe it was the case that the team in the final relegation place had to play-off against the three teams from the division below, so there is precedence for this proposal from the past.
 
CottyImp said:
When the current Play-offs first began, from memory (please correct me if I am wrong) I believe it was the case that the team in the final relegation place had to play-off against the three teams from the division below,
Click to expand...

Yes, that was the arrangement in the first two seasons of the play-offs in the late 1980s. Only one team managed to avoid relegation during that time - Charlton from the First Division in 1987 who thereby denied promotion to Leeds who had finished fourth in the Second Division.
 
Merthyr Imp said:
Yes, that was the arrangement in the first two seasons of the play-offs in the late 1980s. Only one team managed to avoid relegation during that time - Charlton from the First Division in 1987 who thereby denied promotion to Leeds who had finished fourth in the Second Division.
Click to expand...

I remember listening to that game on the radio. Extra time, Leeds scored first then Charlton came back with two goals in the last 7 minutes to stay up.
 
I'm sorry, but as has been stated already, what the hell has the last three years got to do with this season? If roles had been reversed between them and AFC Wimbledon, and Rovers found themselves in 20th spot, three points ahead of the Wombles, with a far better goal margin, would they be dredging up all manner of theories like this? Of course they wouldn't.

Tranmere may have won their last three games of the actual season to suggest they were 'on a roll', but they hadn't won any of their previous 11 league matches before that mini run. Just relegate them and be done with it.
 
The Imposter said:
I'm sorry, but as has been stated already, what the hell has the last three years got to do with this season? If roles had been reversed between them and AFC Wimbledon, and Rovers found themselves in 20th spot, three points ahead of the Wombles, with a far better goal margin, would they be dredging up all manner of theories like this? Of course they wouldn't.

Tranmere may have won their last three games of the actual season to suggest they were 'on a roll', but they hadn't won any of their previous 11 league matches before that mini run. Just relegate them and be done with it.
Click to expand...

The thing that stands out for me with the ‘Last three seasons’ formula is that of course Tranmere had two very good seasons, but they were in lower divisions. How can that be an even playing field compared to teams that were in higher leagues.
I would even query some of their suggestions too, for example saying the National League has a ‘Clear leader’. Barrow are four points clear with nine games to play....that is not ‘Clear’ lead at all and could be easily pegged back.
As already stated, self interest at its finest. I hope they are told to do one by the EFL, however, I won’t be holding my breath on that.
 
Last edited:
The Imposter said:
I'm sorry, but as has been stated already, what the hell has the last three years got to do with this season? If roles had been reversed between them and AFC Wimbledon, and Rovers found themselves in 20th spot, three points ahead of the Wombles, with a far better goal margin, would they be dredging up all manner of theories like this? Of course they wouldn't.

Tranmere may have won their last three games of the actual season to suggest they were 'on a roll', but they hadn't won any of their previous 11 league matches before that mini run. Just relegate them and be done with it.
Click to expand...

I realise that what happened in previous seasons can not directly affect this season, but evidence does suggest that there are a number of examples where teams who were in relegation places at the same 8,9,10 games left in a season did not finish in the relegation zone at the end of the regular season. There are several statistical ways to analyse past evidence. Tranmere have probably picked a sample size of 3 seasons because it does put them on the positive side of the debate. I have to say with my statistical head on, 3 years is a rather small sample and the chance of error in their "error margin" is increased consequently. Perhaps 10 years would have been a better sample size.

We all know anything can happen in football and I guess we have to thank our lucky stars that we are not caught up in the controversy. But surely I'm not the only person out here who can see the grave floor in the EFL's over-simplified PPG suggestion to decide the outcome of the season, not taking into account the split between home games and away and the calibre of team still to be faced??

We don't have crystal balls, but we do have statistics. It's a case of settling on a statistical set that is as reasonable as possible to all concerned in deciding this perilous outcome. I for one would not want to see the prisoner sent down on an unsafe conviction, which, dare I say, feels like a close analogy of what is being proposed by the EFL.
 
CottyImp said:
I realise that what happened in previous seasons can not directly affect this season, but evidence does suggest that there are a number of examples where teams who were in relegation places at the same 8,9,10 games left in a season did not finish in the relegation zone at the end of the regular season. There are several statistical ways to analyse past evidence. Tranmere have probably picked a sample size of 3 seasons because it does put them on the positive side of the debate. I have to say with my statistical head on, 3 years is a rather small sample and the chance of error in their "error margin" is increased consequently. Perhaps 10 years would have been a better sample size.

We all know anything can happen in football and I guess we have to thank our lucky stars that we are not caught up in the controversy. But surely I'm not the only person out here who can see the grave floor in the EFL's over-simplified PPG suggestion to decide the outcome of the season, not taking into account the split between home games and away and the calibre of team still to be faced??

We don't have crystal balls, but we do have statistics. It's a case of settling on a statistical set that is as reasonable as possible to all concerned in deciding this perilous outcome. I for one would not want to see the prisoner sent down on an unsafe conviction, which, dare I say, feels like a close analogy of what is being proposed by the EFL.
Click to expand...

You must be really into numbers and statistics! :grinning:
 
The proposal brings too much messing about and complicates it further IMO. Plus, it also changes the structure of a season (only two down, 25 in L1 next season and expanded PO's) on top of not completing all games.

The main problem is that it's further self interest, as you'd expect it to be. They're trying to fit a system around an end goal.

I'm not really comfortable with the assertions from their end that they'd have stayed up. They were still up against it and arguably had the hardest run of fixtures out of them and the 4 or 5 above them anyway. Unless they had a 9 goal swing, the game in hand is pretty irrelevant as well.
 
Luke Imp said:
The proposal brings too much messing about and complicates it further IMO. Plus, it also changes the structure of a season (only two down, 25 in L1 next season and expanded PO's) on top of not completing all games.

The main problem is that it's further self interest, as you'd expect it to be. They're trying to fit a system around an end goal.

I'm not really comfortable with the assertions from their end that they'd have stayed up. They were still up against it and arguably had the hardest run of fixtures out of them and the 4 or 5 above them anyway. Unless they had a 9 goal swing, the game in hand is pretty irrelevant as well.
Click to expand...

I agree about their run in and that was really my point about the number of home/away games left to play and the calibre of team to be faced. But that really is crystal ball gazing in the absence of a complicated enough statistical hypothesis to test all those variables against (your first point, too complicated !)

Your second point about the League structure, that is why I say if the Tranmere proposal was to be accepted, it should be under the circumstance that they don't just escape relegation, but must take part in the play-offs to provide a proper footballing outcome. As for the expansion of the play-offs, turn it into a one leg affair along the lines of the National League. They have six teams involved, but only play 5 fixtures, the same number of fixtures played in EFL play-offs with only 4 teams. So Play-offs could be expanded without a huge additional game load. E.g 8 teams would require 7 fixtures across three dates in one leg affairs. This could be done in just a 14 day window.

Third point, self interest, absolutely. 3 seasons is too small a sample in my opinion. They have made it big enough to suit their purposes, but not for it to be a thoroughly dependable statistical analysis with enough data to back their assertions reliably.

Apologies, long winded answer to say I mostly agree with what you are saying !
 
CottyImp said:
You know I am my friend :- )=
Click to expand...

CottyImp, you are just what I am looking for. I would really appreciate your views on my observations of the methodology used by Tranmere.

Apologies in advance for the lack of brevity.

A superficial look at the results would say to me there are too many teams qualifying for the play-offs - 12 in the Championship (from Leeds on 71 points down to Derby on 51 points), 8 in L1 and 9 in L2. Also, it doesn't appear to give those teams in the relegation area sufficient margin of error.

Intuitively it doesn't seem right that you can use the same numbers to calculate a margin for error for teams at the top and those at the bottom. The margin of error on the upside for teams that are getting more than 2 points per game is far less than the margin for error of teams at the bottom who might be earning less than a point a game and the opposite for the margin of error on the downside.

So looking at the numbers in a bit more detail you can see how different the margin of error is for teams in the top half and those in the bottom. There are 8 teams in the top half of all 9 tables used that have a margin of error exceeding 10%, only 1 to the upside but 7 to the downside. The bottom half of those tables however results in a completely different picture with 27 teams having a margin of error above 10% but with a significant 20 to the upside and 7 to the downside.

You can also see individual "% error per games" of bottom half teams that would when applied to top teams result in those teams accruing more than 3 points a game over their remaining fixtures, an obviously impossible scenario.

To confirm those observations I used just the three Championship tables to calculate the "% error per game remaining" separately for the top of the table and for the bottom half. My calculations show that the difference between the two halves is as much as 125% to the upside and 24% to the downside.

Its difficult on that basis to believe that the methodology used arrives at an accurate calculation of the margin of error but interested in any views on the above.

In addition their calculations did not take into account that Birmingham City had a 9 point deduction in 2018/19 which overstates the overall "% per game remaining" downside figure used by 3%.
 
You must log in or register to reply here.