Points needed for automatic promotion

[Dimmu]

I thought a little bit of statistical comparisons could be an interesting read. Maybe some statistical knowledge might even calibrate one's expectations.

Here are the points per game averages needed for automatic promotion and playoff spot in recent seasons:

Season      19/20   20/21    21/22    22/23    23/24 so far
top 2          1.77       1.89     1.96       2.13        2.05
playoff       1.69       1.61      1.8         1.67        1.89

This season, we are averaging 1.89.

To get to the average 2.14, which should mean automatic promotion, we would need 62 (61.84) points from the last 27 games. It could be achieved for example with 19 wins, 5 draws, and 3 losses = 98 points.

To get to the average of 2.05, which could be the amount needed for automatics, we would need another 59 (58,3) points. 18 wins, 3 draws, and 6 losses = 93 points.

If we keep getting the points we have so far this season, we will be heading towards another 16 wins, 4 draws, and 7 losses. Overall that would mean 88 points which doesn't sound bad at all, but this season it might not be enough.

People have been saying that the league is poor this season. The stats support the idea as the top 6 teams average more points than usual. With the number of points we have now, we would've been in the top two 20-21 season. Poor level is a double-edged sword though as the top teams don't drop many points. Poor level means unforgiving at the same time.

There is not much to conclude from this except the fact the draws are not enough. We need to go for a win in every game, even if that means occasional loss.

I felt these stats are not completed yet. I had to take a look into xG and xGA. You all wanted this, right? So, here's xGD (expected goal difference) lately.

Season 19/20    20/21    21/22    22/23    23/24
2nd       0.47      0.35      0.77      0.72       0.54
6th        0.21      0.40      0.34      0.24       0.23

We have scored 1.79 per game, while our xG has been 1.51.
We have conceded 0.84 per game when our xGA has been 1.11.

Therefore, we are scoring 0.28 more than expected and conceding 0.27 more than we should, which is a good sign. It also means that we are heading towards xGD 0.4 which usually has meant playoff place but is not enough for automatics.

Let's take a look at other candidates for promotion. If xG doesn't change much towards the end of the season, Peterborough should do well and Stevenage should drop from their current position. Let's see how it goes.

What surprised me most is how well such a robust method as xG, xGA, and xD creates a valid linear pattern over the whole season.

I'm expecting a ridiculous amount of xG bashing even though I'm more interested in what people think of our chance of getting around 95 points this season.

[MackworthRamIsGod]

It's actually good to see that over the last 10 games we are averaging 2.2 points a game.

Carry on with that type of a return and we finish the size mid 90 points.

All eyes on Jan and Feb.

[RadioactiveWaste]

Interesting post, and without reading too deeply into the stats provided it would indicate that we are somewhere where we should be in terms of the top six, and hypothetically if maintained something like our form to date this season we'll be around the top six at the end of the season.

There's always the hope that the team kick on and get better and better now they've figured out what they're doing mostly,bathe likely reality that players will lose form, we'll get some injuries and there'll be games we lose when we shouldn't and games we win when we shouldn't.

Still, blind faith and delusions will see us through regardless. E I E I E I Oh

[Crewton]

46 minutes ago, Dimmu said:

What surprised me most is how well such a robust method as xG, xGA, and xD creates a valid linear pattern over the whole season.

I think the phrase I've emboldened is key : most statistical analysis becomes more accurate and therefore more worthwhile as the number of datasets increase. For individual games, they can be a blunt instrument compared to other statistics and visual evidence.

The overall analysis is interesting though - thanks! 

[Ramarena]

1 hour ago, MackworthRamIsGod said:

It's actually good to see that over the last 10 games we are averaging 2.2 points a game.

Carry on with that type of a return and we finish the size mid 90 points.

All eyes on Jan and Feb.

Feb and March are the ones. It’s when we’ve always fallen apart in recent years. 

Hopefully this season we can keep going and not succumb to the struggles of season gone by!

[DavesaRam]

4 minutes ago, Ramarena said:

Feb and March are the ones. It’s when we’ve always fallen apart in recent years. 

Hopefully this season we can keep going and not succumb to the struggles of season gone by!

Yes. We need to realise we aren’t the Derby County of 22/23, or of 21/22, or of 20/21 et al. We are a different team, we are a  different squad so we aren’t any of the teams that fell away in February and March, so it ain’t going to happen.

And if I poke my eyes out I’ll be a blind-faither!

[Matchday Fred]

It’s a dull cliche but focus should be on the next game. Do not distract the players from tomorrow’s job, keep it simple but effective. Celebrate wins and reinforce strengths. Projecting forward and obsessing with fixtures and scoreboards is usually a mugs game. Focus on current form and what we can do better.  What we need this year is to start beating the sides in the top 6, setting out to win every game especially away from home, getting an early goal, being the side on top in the last 10 mins, not settling for a draw because legs are tired.
 

There for the taking if we stay fit, keep our s*** together. 

Edited December 12, 2023 by Matchday Fred

[RodleyRam]

Thanks for the analysis on this.

I think most of us would say we started slowly and are now accelerating. If we can maintain the kind of form in more recent months, then we should be looking to improve on our average PPG so far. However, whilst many have said it's a weaker league, the gritty relentlessness of L1 will always throw up disappointing results and dips in form. We have to bounce back from disappointment better than last season.

I stand by my pre-season prediction of a play off spot, which I think would be a decent season. I really don't buy into the 'anything less than promotion is a failure' narrative. Anything beyond playoffs is within our capability but I would see as a happy bonus.

[Chopper]

Although looking at figures usually bring me out in a rash, nice work Dimmu. 

4 hours ago, Dimmu said:

If we keep getting the points we have so far this season, we will be heading towards another 16 wins, 4 draws, and 7 losses. Overall that would mean 88 points which doesn't sound bad at all, but this season it might not be enough.

I have a feeling in my gut (which may or may not be because Mrs. Chopper does all my meat, medium rare, including chicken) that our points per game for the second half of the season will be better than the first half, and also that the points needed for automatics will be lower. Some bookies apparently, somehow, have 8 teams as odds on for top 6 (no me neither), but it is indicative of the lack of standout teams and it's those teams that mean crazy figures for teams in the autos.

I might, just might be starting to feel the butterflies of optimism.

[Dimmu]

10 minutes ago, Chopper said:

Although looking at figures usually bring me out in a rash, nice work Dimmu. 

I have a feeling in my gut (which may or may not be because Mrs. Chopper does all my meat, medium rare, including chicken) that our points per game for the second half of the season will be better than the first half, and also that the points needed for automatics will be lower. Some bookies apparently, somehow, have 8 teams as odds on for top 6 (no me neither), but it is indicative of the lack of standout teams and it's those teams that mean crazy figures for teams in the autos.

I might, just might be starting to feel the butterflies of optimism.

I think we'll do well in the other half of the season as well. This is based on blind hope 😄 

Actually, I planned to check what's been the quarterly average point return of Warne's teams during his promotion seasons but analyzing takes too much time as I need also to create some sort of database. As Warne is known for his fitness-based approach, there could be some interesting insights available. 

Maybe I'll do it later if people are interested in reading such an analysis.

[Allen]

Think so much will depend on squad depth as injuries, suspensions and fatigue take their toll.

Warne knows we need reinforcements in January and for me, we need 3 new players.

My fear is that Cashin and Bird leave which would mean 5 new players needed. To bring in 5 quality players in a notoriously difficult window would be some achievement. Planning seems to have been ongoing for some time so hopefully we can do it.

[Sparkle]

Whilst we will strive for as many points as we can the other teams at the top seem to rarely drop many points but as we know that can all go pear shaped and in some style 

[Gee SCREAMER !!]

41 minutes ago, Allen said:

Think so much will depend on squad depth as injuries, suspensions and fatigue take their toll.

Warne knows we need reinforcements in January and for me, we need 3 new players.

My fear is that Cashin and Bird leave which would mean 5 new players needed. To bring in 5 quality players in a notoriously difficult window would be some achievement. Planning seems to have been ongoing for some time so hopefully we can do it.

I think we have 11 players out of contract next summer and 2 more in January.  Not so much of a problem if we go up as we'll need to upgrade in the squad .  Quite a problem if we don't as a whole scale at League 1 when your going into year 3 would be tricky. 

[Ghost of Clough]

Wgat do we need? A new club record. The most points we gained in a season was 85 in 13/14.

We also gained 63 points in 55/56 during our 1st spell in the 3rd division, back when it was 2 points for a win (equivalent to 91 points)

[ram59]

6 hours ago, Dimmu said:

I thought a little bit of statistical comparisons could be an interesting read. Maybe some statistical knowledge might even calibrate one's expectations.

Here are the points per game averages needed for automatic promotion and playoff spot in recent seasons:

Season      19/20   20/21    21/22    22/23    23/24 so far
top 2          1.77       1.89     1.96       2.13        2.05
playoff       1.69       1.61      1.8         1.67        1.89

This season, we are averaging 1.89.

To get to the average 2.14, which should mean automatic promotion, we would need 62 (61.84) points from the last 27 games. It could be achieved for example with 19 wins, 5 draws, and 3 losses = 98 points.

To get to the average of 2.05, which could be the amount needed for automatics, we would need another 59 (58,3) points. 18 wins, 3 draws, and 6 losses = 93 points.

If we keep getting the points we have so far this season, we will be heading towards another 16 wins, 4 draws, and 7 losses. Overall that would mean 88 points which doesn't sound bad at all, but this season it might not be enough.

People have been saying that the league is poor this season. The stats support the idea as the top 6 teams average more points than usual. With the number of points we have now, we would've been in the top two 20-21 season. Poor level is a double-edged sword though as the top teams don't drop many points. Poor level means unforgiving at the same time.

There is not much to conclude from this except the fact the draws are not enough. We need to go for a win in every game, even if that means occasional loss.

I felt these stats are not completed yet. I had to take a look into xG and xGA. You all wanted this, right? So, here's xGD (expected goal difference) lately.

Season 19/20    20/21    21/22    22/23    23/24
2nd       0.47      0.35      0.77      0.72       0.54
6th        0.21      0.40      0.34      0.24       0.23

We have scored 1.79 per game, while our xG has been 1.51.
We have conceded 0.84 per game when our xGA has been 1.11.

Therefore, we are scoring 0.28 more than expected and conceding 0.27 more than we should, which is a good sign. It also means that we are heading towards xGD 0.4 which usually has meant playoff place but is not enough for automatics.

Let's take a look at other candidates for promotion. If xG doesn't change much towards the end of the season, Peterborough should do well and Stevenage should drop from their current position. Let's see how it goes.

What surprised me most is how well such a robust method as xG, xGA, and xD creates a valid linear pattern over the whole season.

I'm expecting a ridiculous amount of xG bashing even though I'm more interested in what people think of our chance of getting around 95 points this season.

Your play off figure of 1.89 for this season's play offs is a touch out, at the time of your post Stevenage's 1.86 was sixth best, but now it's Oxford's 1.85 ppg. Your point is still valid about it being the highest figure in recent years though. 

[Ram@Lincoln]

I've always ran a rule of thumb that 30 wins in the league season will get you promoted. Likewise with the goal difference, concede less than a goal a game average and score around the 80 mark and you'll be right about there.

Therefore, we need 19 wins out of 27 games, 46 more goals scored and can afford 30 goals conceded (obviously less than be ideal). With 19 wins added to what we have now, even allowing for losses in the other games that's 93 points which gets us promoted 7 out the last 9 seasons.

[kevinhectoring]

14 hours ago, Dimmu said:

I thought a little bit of statistical comparisons could be an interesting read. Maybe some statistical knowledge might even calibrate one's expectations.

Here are the points per game averages needed for automatic promotion and playoff spot in recent seasons:

Season      19/20   20/21    21/22    22/23    23/24 so far
top 2          1.77       1.89     1.96       2.13        2.05
playoff       1.69       1.61      1.8         1.67        1.89

This season, we are averaging 1.89.

To get to the average 2.14, which should mean automatic promotion, we would need 62 (61.84) points from the last 27 games. It could be achieved for example with 19 wins, 5 draws, and 3 losses = 98 points.

To get to the average of 2.05, which could be the amount needed for automatics, we would need another 59 (58,3) points. 18 wins, 3 draws, and 6 losses = 93 points.

If we keep getting the points we have so far this season, we will be heading towards another 16 wins, 4 draws, and 7 losses. Overall that would mean 88 points which doesn't sound bad at all, but this season it might not be enough.

People have been saying that the league is poor this season. The stats support the idea as the top 6 teams average more points than usual. With the number of points we have now, we would've been in the top two 20-21 season. Poor level is a double-edged sword though as the top teams don't drop many points. Poor level means unforgiving at the same time.

There is not much to conclude from this except the fact the draws are not enough. We need to go for a win in every game, even if that means occasional loss.

I felt these stats are not completed yet. I had to take a look into xG and xGA. You all wanted this, right? So, here's xGD (expected goal difference) lately.

Season 19/20    20/21    21/22    22/23    23/24
2nd       0.47      0.35      0.77      0.72       0.54
6th        0.21      0.40      0.34      0.24       0.23

We have scored 1.79 per game, while our xG has been 1.51.
We have conceded 0.84 per game when our xGA has been 1.11.

Therefore, we are scoring 0.28 more than expected and conceding 0.27 more than we should, which is a good sign. It also means that we are heading towards xGD 0.4 which usually has meant playoff place but is not enough for automatics.

Let's take a look at other candidates for promotion. If xG doesn't change much towards the end of the season, Peterborough should do well and Stevenage should drop from their current position. Let's see how it goes.

What surprised me most is how well such a robust method as xG, xGA, and xD creates a valid linear pattern over the whole season.

I'm expecting a ridiculous amount of xG bashing even though I'm more interested in what people think of our chance of getting around 95 points this season.

The point you make about playing for wins not draws is an interesting one. I think you have to have a clear strategy on this and need to be willing to change course in-play. I also think you need to factor in who you’re playing, H/A,  and where the oppo sits in the table. 
 

we have had a really good run points wise. It flatters our form (probably reflecting the quality of the league). Sometimes a team on a winning streak can tweak and play itself into better form. If we can do this I think we have a good chance of automatics especially if we strengthen in Jan
 

What I fear is that, if we fail to make the automatics, we don’t have the game management skills to win the play offs even though we have a strong squad 

[Dimmu]

10 hours ago, ram59 said:

Your play off figure of 1.89 for this season's play offs is a touch out, at the time of your post Stevenage's 1.86 was sixth best, but now it's Oxford's 1.85 ppg. Your point is still valid about it being the highest figure in recent years though. 

Thanks for the correction, I must've been thinking of Derby's ppg when typing. Unfortunately, I can't correct it to the original post anymore. 😕 

Edited December 13, 2023 by Dimmu

[FindernRam]

4 hours ago, kevinhectoring said:

The point you make about playing for wins not draws is an interesting one.

Isn't this why 3 points for a win was introduced: to encourage attacking play, particularly away from home.

With so many games still to play and form (particularly in the lower leagues) being so variable I'm not overly excited by targets as such, only that we play to win our games and stay in touch (at least) with the top positions.

I was a statistician so am fascinated by the ways Football is analysed to death these days. Stats 101 states that predictions are more likely prove right with more data points. BUT at the individual game level the numbers are so low that any prediction of any characteristic has a very low probability of being correct. Be that goals, throw-ins wins draws.

Ask yourself why betting organisations are helpfully pushing us to use these numbers. It encourages us to believe we can do a better job of weighing the odds than they can, so we play more. Thing is they have masses of data points to set the odds and can happily (from their point of view) lose some because overall they will come out ahead.

Its not quite "snake oil" but probably better than Horace Batchelors Infra draw system for K E N Y S H A M Bristol.

[Stive Pesley]

2 minutes ago, FindernRam said:

I was a statistician so am fascinated by the ways Football is analysed to death these days. Stats 101 states that predictions are more likely prove right with more data points. BUT at the individual game level the numbers are so low that any prediction of any characteristic has a very low probability of being correct. Be that goals, throw-ins wins draws.

Indeed - game stats are never going to be reliably predictable because a team has 11 moving parts (or in our case 10 plus Hourihane)