Topic: The Alternative Premier League Table 2015-16

[Hij]

not sure whether the league is tougher or everyone is still finding their feet.

I was wondering this. Both I suspect. Surely with the TV money now some of the smaller teams will become better? I guess we'll see what the final points totals are, probably the same as usual, but wonder if we'll see a lower winners total moving forwards.

[RooiBefok]

The thirteen teams that “qualify” for inclusion in Prof’s APLT model.

[McrRed]

The thirteen teams that “qualify” for inclusion in Prof’s APLT model.

And yet  it'll be the same 6 that will feature later in the season. Finances function as a sort of meta-Alternative Premier League.

[Prof]

The thirteen teams that “qualify” for inclusion in Prof’s APLT model.

Nice work

Thanks for sharing

[sminp]

The thirteen teams that “qualify” for inclusion in Prof’s APLT model.

Nice addition to the thread. I wouldn't mind seeing this for all 20 teams then we can see who's likely to be relegated too if you can be arsed.

[Prof]

Nice addition to the thread. I wouldn't mind seeing this for all 20 teams then we can see who's likely to be relegated too if you can be arsed.

The further you go down the league, the less valid this approach becomes.  I might be worth looking at a model for lower half of the league teams.  I think we discussed it a few seasons ago.

[Elmo!]

The further you go down the league, the less valid this approach becomes.  I might be worth looking at a model for lower half of the league teams.  I think we discussed it a few seasons ago.

Surely an Everton fan has looked into this? 

[Prof]

The further you go down the league, the less valid this approach becomes.  I might be worth looking at a model for lower half of the league teams.  I think we discussed it a few seasons ago.

I think a par of 47 to finish 11th (win the bottom half of the league  ) would be a decent benchmark.  Looking at it, I'd probably look at;

7 home wins
12 home draws

(33 points from home)

14 away draws

(14 points from away)

I'd then factor in a 0.25 point drop per match for the relagation line (-9.5 or 37.5 points by the end of the season)

[SP]

How about a relegation one?. 10 easiest always have draw as a par, 10 easiest homes have par 3. That is 40 points, the traditional safety point. If you are positive on that par, you are on course to stay up.

[Prof]

How about a relegation one?. 10 easiest always have draw as a par, 10 easiest homes have par 3. That is 40 points, the traditional safety point. If you are positive on that par, you are on course to stay up.

That's an option.

Teams that miss the drop look like they get about 10 wins a season, so that fits.  Makes sense to say they'll be all at home, even if we know this unlikely in practice.

The share of the draws maybe needs to be weighted more to home matches though.  Draw away at 11 above draw at home to 10 for example is a difficult one to justify.

[riismeister]

That's an option.

Teams that miss the drop look like they get about 10 wins a season, so that fits.  Makes sense to say they'll be all at home, even if we know this unlikely in practice.

The share of the draws maybe needs to be weighted more to home matches though.  Draw away at 11 above draw at home to 10 for example is a difficult one to justify.

Perhaps something like
10 wins at home to the bottom 10 teams previous season = 30 points
5 draws away to the bottom 5 = 5 points
5 draws at home to the 5 teams above the bottom 10 = 5
= total of 40 points

[zabadoh]

Or some could take a look at the records of the teams that finished in 17th place over, say 20 seasons, then compute averages or means in WLDs

[xRedmanLFCx]

The current graph only has us and city above the top 4 line. That has cheered me up a lot! Thanks Prof!

[RooiBefok]

Nice addition to the thread. I wouldn't mind seeing this for all 20 teams then we can see who's likely to be relegated too if you can be arsed.

[Sudden Death Draft Loser]

The current graph only has us and city above the top 4 line. That has cheered me up a lot! Thanks Prof!

How?

Clearly i really don't understand the alternative PL table nonsense.

[Brian Blessed]

Have you tried?

[Hij]

Must be some effort that. I'll have to throw you some beer money at the end of the season if you keep it up!

[Nessy76]

Nice work. I'm sure you have thought of this, but for teams outside the top 13 last season, there should still only be twelve par one matches.

[RooiBefok]

Must be some effort that. I'll have to throw you some beer money at the end of the season if you keep it up!

 

[RooiBefok]

Nice work. I'm sure you have thought of this, but for teams outside the top 13 last season, there should still only be twelve par one matches.

Correct.  Hence why I initially only expanded Prof’s table to include the teams finishing in log positions 1 to 13 (inclusive) from last year’s Premier League Table.

Having been requested to include all 20 teams I used the following rationale.

Prof’s premise has always been that in order for a team to tally 90 log points in the season, they need to win their 19 Home matches (19 x 3 = 57) plus the seven easiest Away matches (7 x 3 = 21) and draw their remaining 12 Away matches (12 x 1 = 12), resulting in a total tally of 90 log points.  Given that Prof has always limited his APLT analysis to the “top” seven teams, the seven easiest Away matches have naturally been those teams finishing in log positions 14 to 20 (inclusive) from last year’s Premier League Table. [3 relegated teams replaced with 3 promoted teams].

However, now that we wish to include the last seven teams from last year in Prof’s APLT analysis the points tally will compute as follows; (19 x 3 = 57) + (6 x 3 = 18) + (13 x 1 = 13) resulting in a total tally of 88 log points.

This is ‘workable’ but if the APLT methodology is confusing to a large number now, it will cause much gnashing of teeth if we were to introduce a total par tally of 90 for thirteen teams and another total par tally of 88 for seven teams.

So, keeping in mind Prof’s founding principle that in order for a team to win the title that team needs to tally 90 log points, I have simply introduced the following caveat to my APLT logic.

In order for a team to tally 90 log points in the season, they need to win their 19 Home matches plus the seven easiest Away matches, (the three promoted teams and the four lowest ranked teams excluding oneself, from the previous season) and draw their remaining Away matches.

The resultant effect being that teams 1 to 12 need to only draw Away to the team that finished in 13th position (West Brom) last year, while teams 14 to 20 need to win against West Brom.

Teams 1 to 13: (19 x 3 = 57) + (7 x 3 = 21) + (12 x 1 = 12) = 90
Teams 14 to 20: (19 x 3 = 57) + (7 x 3 = 21) + (12 x 1 = 12) = 90

We are now back to square one and all 20 teams can be included.

I know…. clear as mud!

[RooiBefok]

I know…. clear as mud!

[Nessy76]

Good work! And that does seem like the most logical solution.

[Franck Le Poof]

Just wanted to say how much I appreciate the work put in here. I've only just come across this thread for the first time; I used to read the title and assume it was a "what if" sort of shindig. Really appreciate the analysis and the different perspective it forces one to examine results through.

Brilliant stuff. Cheers.

[Prof]

It'll be interesting looking at the lower ranked teams using this model, but I don't think it's as good an indicator the further down the league you go.

It's a model based on winning the league, so by definition, there aren't many matches where dropping points is 'acceptable'.

Teams looking to avoid the drop will have lots of matches where they 'have nothing to lose' or where a draw is 'dream land'.  Their bread and butter should be home matches particularly against teams in the bottom half.

Using the aplt for teams at the bottom will treat home matches against the best and worst teams the same.  In reality, winning 10 home matches is almost enough to keep you up on their own.  Draws against top half teams will be good results.

If people are interested, I'll try to find time to develop an APLT Relegation Edition.  I'm torn between using 40 points as par or 47 as par.  If I was working with a club, I wouldn't aim for 40, but 47.

[zabadoh]

For the last 20 seasons, 17th place:

Points:  Average 38.35, Mean (exclude top bottom 10%) 38.28, Mode 38
GD:  Average -20.25, Mean -20.28, Mode -22

If anything, the trend for points for survival has been averaging lower at 37.2 pts these last 10 seasons down from 39.5 pts the previous 10 seasons.

Here's a Google spreadsheet that I've put together about 17th place:

https://docs.google.com/spreadsheets/d/1unaehMnE3ZWAbcjt2PAsMhEkc_lvC-eEY9jMBjszq-Q/edit?usp=sharing

[Ycuzz]

The thirteen teams that “qualify” for inclusion in Prof’s APLT model.

Nice swan! Thanks.

Now for the entire league

[Peabee]

Perhaps something like
10 wins at home to the bottom 10 teams previous season = 30 points
5 draws away to the bottom 5 = 5 points
5 draws at home to the 5 teams above the bottom 10 = 5
= total of 40 points

What's that based on? Statistical analysis of previous seasons?  Or plucked from the air? 

[Nessy76]

What's that based on? Statistical analysis of previous seasons?  Or plucked from the air? 

I would expect any serious analysis would soon show that it's almost impossible to predict which games teams will pick up points in at the lower end of the table. For title contenders, it's a lot easier because they will win the majority of matches, but getting forty points from 38 games has a far wider margin of error.

The APLT does at least provide a blueprint showing which games should be easier to win, although I doubt that many sides that finished 17th would come close to following that method.

It's an interesting project and might be of some interest but it is unlikely to ever match the elegance of the APLT as we know it.

[rickardinho1]

What an enormous chance the United game is to open a gap. A win puts us on +1 and drops United to -6.
 
Also Chelsea and City have tough games away at Everton and Palace respectively (both par 1's), so it's not unfeasible that the table could look like this by the end of the weekend (Arsenal should win vs Stoke at home):

City +3
Liverpool +1
Arsenal -3
Chelsea -5
United -6

It's probably unlikely that both Chelsea and City lose, but regardless, opening a 7 point gap on United to effectively push them to 5th place in the top-4-chasing mini-league would be massive, even if the gap in the real table would only be 3 points.

[zabadoh]

I would expect the relegation endangered clubs to take points off of each other.

No point speculating, if someone gets some time, do an actual study.

[Nessy76]

The APLT weekend preview is here.

With the nonsense of the international break now behind us, and with nothing of any interest having happened in football over the last two weeks as a consequence, we're all desperate for some football as the nascent Premier League season gets back on track.

And what a feast the APLT serves up this weekend.

First off, Cheeky Champions Chelsea chase chances to choke the chaos that has characterised their charge to the championship this season. The Russian crime-ring laundry service visit Everton in a par 1 game without the services of their talismanic keeper Courtois.
The Toffees have started solidly this season, and much will be made of their decision to retain the services of John Stones, who will now be employed to keep his admirers off the score sheet. Chelsea have struggled in their fixtures so far, with just a solitary win away to West Brom. On form, Everton have this in the bag. Their only defeat has come against Manchester City, and they already look a more convincing outfit than the one that lost just five home games last season.

It's a big Par 1 for Chelsea, who will need to rediscover their usual form and make some progress back towards the 0 line of the APLT.

Chelsea win: CHE -2
Draw: CHE -4
Everton win: CHE -5

Arsenal entertain Stoke in a Par 3 game at the Emirates, with the Potters yet to win a game this season, despite boasting some serious quality options in attack, added to their traditionally miserly defence. The balance has yet to be struck, it seems, and with just five away wins under their belts last season, the Emirates looks an unlikely place for that to happen, despite Arsenal's own slow start and the Gunners will be looking for their first home win before their fans start the annual "should we sack Wenger" debate early.

Arsenal win: ARS -3
Draw: ARS -5
Stoke win: ARS -6

APLT leaders Manchester City, meanwhile, travel to Crystal Palace, hoping to preserve their 100% start to the campaign. It's first v second in the trad table, and conventional wisdom tells us Palace will only be moving in one direction from there. A credible top half finish for Palace last term saw them lose just six games on the road throughout last season, though, and coming off the back of a rare away victory at Stamford Bridge, Pardew's lads should not be overawed by the occasion in front of their own fans. It's a winnable Par 1 game on paper, but the APLT tells us City will do well to come away with the draw.

Manchester City win: MNC +6
Draw: MNC +4
Palace win: MNC +3

Southampton visit West Brom in a bid to claw back APLT respectability. It's the softest Par 1 on the fixture list, and the Saints will want very badly to improve on their -7 APLT points with a win. The Baggies were beaten on eight separate occasions at home last season, and have yet to win a home league match this term, although with Chelsea and Manchester City the only visitors so far, that may be less telling than it looks.

Southampton win: SOT -5
Draw: SOT -7
West Brom win: SOT -8

The biggest game of the weekend is in Manchester, where Liverpool make the short journey to visit fierce rivals Manchester United. There's little that hasn't already been said about this fixture, but Brendan Rodgers' team will be looking at the APLT table with a smile, knowing that this is a fragile Man United who could well be there for the taking. The Red Devils lost just three games at the Theatre of Tantrums last season, and while their last home appearance saw them grind out a draw with Newcastle, this is a different game entirely. Liverpool's bright start was cut short with a frustrating home loss to West Ham United before the break, and the reds will be hoping for a performance closer to that they gave against Arsenal. It's a par 1 for Liverpool, a par 3 for Man Utd, and Liverpool fans probably won't be content with the knowledge that they will remain above their horrible foes in the APLT regardless of the result, when a win could open a seven point gap.

Man Utd win: MNU -3 LFC -2
Draw: MNU -5 LFC -1
Liverpool win: MNU -6 LFC +1

And finally, bottom-lurking Spurs take the long road North to Sunderland on Sunday, like a bunch of rapists from King's Landing, sent off to take the black, join the Night's Watch and live short and violent lives of celibacy. (Harry Kane never looks like scoring again anyway.) At least the Wildlings of the SPL aren't expected to invade this weekend.
Tottenham's deficiencies have been much pored over in recent weeks, but it's always funny to do so again. The Spurs sit 16th in the trad table with successive draws against Leicester, Stoke and Everton in their last three league outings representing their only points of the campaign. The fact Sunderland could only win four games last season, and now occupy the bottom spot of the trad table should raise the hopes of Tottenham supporters, which is always funny later if they are dashed again cruely.

This, of course, is a Par 3 game, and if Spurs don't win it, harsh questions will certainly be asked on Match of the Day 2. Probably by Garth. He loves asking those hard questions. It gives him an excuse to wag his finger and try to sound ominous. And given his beloved Tottenham could end the weekend on -10 APLT points, he might well have a point for once. (Even if he is unlikely to make it in any coherent manner.)

Tottenham win: SPU -7
Draw: SPU -9
Sunderland win: SPU -10

[SamAteTheRedAcid]

 

And finally, bottom-lurking Spurs take the long road North to Sunderland on Sunday, like a bunch of rapists from King's Landing, sent off to take the black, join the Night's Watch and live short and violent lives of celibacy. (Harry Kane never looks like scoring again anyway.) At least the Wildlings of the SPL aren't expected to invade this weekend.

[Peabee]

I would expect any serious analysis would soon show that it's almost impossible to predict which games teams will pick up points in at the lower end of the table. For title contenders, it's a lot easier because they will win the majority of matches, but getting forty points from 38 games has a far wider margin of error.

The APLT does at least provide a blueprint showing which games should be easier to win, although I doubt that many sides that finished 17th would come close to following that method.

It's an interesting project and might be of some interest but it is unlikely to ever match the elegance of the APLT as we know it.

I'm a Mathematican, we're pedants by nature, hence the  .

[riismeister]

What's that based on? Statistical analysis of previous seasons?  Or plucked from the air? 

Plucked from the air 

But I don't think 40 points is a bad cut off point for safety.

[Nessy76]

Oof, big old kick in the swingers for Mourinho there. Chelsea all over the place and Everton making a strong case for their return to the APLT.

[Prof]

Oof, big old kick in the swingers for Mourinho there. Chelsea all over the place and Everton making a strong case for their return to the APLT.

Only -1

They've had three par one matches in the first five.  They won't look as bad here as the real table.  What's interesting is how much the real table affects confidence etc and that of opponents.

[McrRed]

The APLT weekend preview is here.

With the nonsense of the international break now behind us, and with nothing of any interest having happened in football over the last two weeks as a consequence, we're all desperate for some football as the nascent Premier League season gets back on track.

And what a feast the APLT serves up this weekend.

First off, Cheeky Champions Chelsea chase chances to choke the chaos that has characterised their charge to the championship this season. The Russian crime-ring laundry service visit Everton in a par 1 game without the services of their talismanic keeper Courtois.
The Toffees have started solidly this season, and much will be made of their decision to retain the services of John Stones, who will now be employed to keep his admirers off the score sheet. Chelsea have struggled in their fixtures so far, with just a solitary win away to West Brom. On form, Everton have this in the bag. Their only defeat has come against Manchester City, and they already look a more convincing outfit than the one that lost just five home games last season.

It's a big Par 1 for Chelsea, who will need to rediscover their usual form and make some progress back towards the 0 line of the APLT.

Chelsea win: CHE -2
Draw: CHE -4
Everton win: CHE -5

Arsenal entertain Stoke in a Par 3 game at the Emirates, with the Potters yet to win a game this season, despite boasting some serious quality options in attack, added to their traditionally miserly defence. The balance has yet to be struck, it seems, and with just five away wins under their belts last season, the Emirates looks an unlikely place for that to happen, despite Arsenal's own slow start and the Gunners will be looking for their first home win before their fans start the annual "should we sack Wenger" debate early.

Arsenal win: ARS -3
Draw: ARS -5
Stoke win: ARS -6

APLT leaders Manchester City, meanwhile, travel to Crystal Palace, hoping to preserve their 100% start to the campaign. It's first v second in the trad table, and conventional wisdom tells us Palace will only be moving in one direction from there. A credible top half finish for Palace last term saw them lose just six games on the road throughout last season, though, and coming off the back of a rare away victory at Stamford Bridge, Pardew's lads should not be overawed by the occasion in front of their own fans. It's a winnable Par 1 game on paper, but the APLT tells us City will do well to come away with the draw.

Manchester City win: MNC +6
Draw: MNC +4
Palace win: MNC +3

Southampton visit West Brom in a bid to claw back APLT respectability. It's the softest Par 1 on the fixture list, and the Saints will want very badly to improve on their -7 APLT points with a win. The Baggies were beaten on eight separate occasions at home last season, and have yet to win a home league match this term, although with Chelsea and Manchester City the only visitors so far, that may be less telling than it looks.

Southampton win: SOT -5
Draw: SOT -7
West Brom win: SOT -8

The biggest game of the weekend is in Manchester, where Liverpool make the short journey to visit fierce rivals Manchester United. There's little that hasn't already been said about this fixture, but Brendan Rodgers' team will be looking at the APLT table with a smile, knowing that this is a fragile Man United who could well be there for the taking. The Red Devils lost just three games at the Theatre of Tantrums last season, and while their last home appearance saw them grind out a draw with Newcastle, this is a different game entirely. Liverpool's bright start was cut short with a frustrating home loss to West Ham United before the break, and the reds will be hoping for a performance closer to that they gave against Arsenal. It's a par 1 for Liverpool, a par 3 for Man Utd, and Liverpool fans probably won't be content with the knowledge that they will remain above their horrible foes in the APLT regardless of the result, when a win could open a seven point gap.

Man Utd win: MNU -3 LFC -2
Draw: MNU -5 LFC -1
Liverpool win: MNU -6 LFC +1

And finally, bottom-lurking Spurs take the long road North to Sunderland on Sunday, like a bunch of rapists from King's Landing, sent off to take the black, join the Night's Watch and live short and violent lives of celibacy. (Harry Kane never looks like scoring again anyway.) At least the Wildlings of the SPL aren't expected to invade this weekend.
Tottenham's deficiencies have been much pored over in recent weeks, but it's always funny to do so again. The Spurs sit 16th in the trad table with successive draws against Leicester, Stoke and Everton in their last three league outings representing their only points of the campaign. The fact Sunderland could only win four games last season, and now occupy the bottom spot of the trad table should raise the hopes of Tottenham supporters, which is always funny later if they are dashed again cruely.

This, of course, is a Par 3 game, and if Spurs don't win it, harsh questions will certainly be asked on Match of the Day 2. Probably by Garth. He loves asking those hard questions. It gives him an excuse to wag his finger and try to sound ominous. And given his beloved Tottenham could end the weekend on -10 APLT points, he might well have a point for once. (Even if he is unlikely to make it in any coherent manner.)

Tottenham win: SPU -7
Draw: SPU -9
Sunderland win: SPU -10

Fantastic write up Nessy [emoji106]

[Prof]

[sminp]

We're still 2nd!  This makes me feel slightly better

[Tumbleweed]

Ssssh haven't you heard? Seasons over