The Adjusted League Table

[Gillespie]

Yes but this is still entirely reflective of who you have played. Man City haven't had the chance to face a team to get points off while Chelsea have, so it has the exact same flaws, ie it takes ages for a pattern to emerge and the early pattern is entirely based on fixtures.

In other words just use your brain to establish who is on form.

Obviously.Any measure of performance must be based upon actual results.Pete's sytem is not meant to be perfect, if it were you would be able now to predict the table in mid May 2010.

What Pete's system does though,is to weight those early season games according to the relative difficulty of the tie so as to give a better guide just from a few results as to the likely table come the season's end.It's a good guide and easier to understand than Daniel Finkelstein's Fink tank.As the season progresses it should converge with the actual league table.

 

[Rood]

The idea of the table is to adjust the points according to the difficulty of the fixtures already played, with the idea being the team at the top is best placed to win the league. The scoring system is as follows:

Win home fixtures (57 pts)
Win away fixtures v bottom 8 (24 pts)
Draw away fixtures v middle 8 (8 pts)
Lose away fixtures v other top 4 (0pts)

Dont fully agree with some of the assumptions behind the scoring system but it is good enough for a basic system - i see that someone else tried to refine it with a higher level of complexity and I suppose you could continue improving it and eventually end up with a specific weighting for every team in the league but probably not worth the bother.

I am glad that other people can be arsed to work this all out as it seems more effort than it is worth - that being said, will be interesting to follow it throughout the season as an experiment - has anyone bothered to apply it to previous seasons and see what it comes out with?

 

[Gillespie]

Dont fully agree with some of the assumptions behind the scoring system but it is good enough for a basic system - i see that someone else tried to refine it with a higher level of complexity and I suppose you could continue improving it and eventually end up with a specific weighting for every team in the league but probably not worth the bother.

I am glad that other people can be arsed to work this all out as it seems more effort than it is worth - that being said, will be interesting to follow it throughout the season as an experiment - has anyone bothered to apply it to previous seasons and see what it comes out with?

Apart from the time needed to manufacture and then having to apply even more complex parameters to the relative measures of performance, I think you would find that a kind of Heisenberg's Uncertainty Principle would come into play i.e. the more you attempt to measure the unmeasurable the less likely you would be of getter any greater accuracy.If that makes sense?

 

[Rood]

Apart from the time needed to manufacture and then having to apply even more complex parameters to the relative measures of performance, I think you would find that a kind of Heisenberg's Uncertainty Principle would come into play i.e. the more you attempt to measure the unmeasurable the less likely you would be of getter any greater accuracy.If that makes sense?

Never heard of Heisenberg but yes it actually does make sense !

 

[MUFCgal]

I am incredibly confused right now.

Me too

 

[Nistelrooy10]

Those of you who listen to Over the Bar will already be familiar with Peter Storey's league table, but it deserves it's own thread. It will be kept updated throughout the season.

The idea of the table is to adjust the points according to the difficulty of the fixtures already played, with the idea being the team at the top is best placed to win the league. The scoring system is as follows:

Win home fixtures (57 pts)
Win away fixtures v bottom 8 (24 pts)
Draw away fixtures v middle 8 (8 pts)
Lose away fixtures v other top 4 (0pts)

I like it. I'm sure SAF has a similar system of his own. It only makes sense to give weight to certain wins and take away from others. The results are a bit worrying, however, we need to take into account is the manner Chelsea came to the top; scoring some late goals skewed that table a little.

 

[theimperialinn]

I think a team should be awarded brownie points for wearing an attractive kit. And if at least one player makes onto Gay Times Sexiest Man of the Year list they should get 6

I thought Bolton's new shirt was a bit snazzy the other day.

 

[surf]

It's a good starting point and Pete touted it quite successfully last season to show that our seemingly poor start was accounted for by having difficult fixtures.

I'd suggest redefining the middle and bottom 8 periodically as form emerges. Also, a home win over a top 4 rival is better than a par.

 

[MikeUpNorth]

I've edited the first post to reflect today's results.

 

[Laphroaig]

I've edited the first post to reflect today's results.

And I've edited the second post. We've climbed above Chelsea in Rowem's system.

 

[Valentino]

Table in a second post is out of date

 

[Laphroaig]

Updated.

Quite a difference between the two tables.

 

[Plechazunga]

Can we have an update please?

 

[Brophs]

It's in each week's podcast thread, no?

cnut.

 

[Brophs]

https://www.redcafe.net/f7/over-bar-episode-12-a-274585/#post7126460

 

[Plechazunga]

Thanks, but there's no need to take that rude tone!

 

[Brophs]

Thanks you patronising cnut

Do you even know what patronising means, little man?

 

[Plechazunga]

I think old Brophs might be losing the plot

 

[Brophs]

Your mum patronises crack-houses

You can't win every argument with that response, you know.



And she thought it was an early house. That's why she was in there.

 

[MikeUpNorth]

Can we have an update please?

Done.

Can we have some more hilarious sketches please?

 

[Plechazunga]

Can I suggest a refinement for next season?

The biggest inaccuracy as I see it is the initial rankings of teams. A side like West Ham can finish high up one season and easily be in the bottom 5 most of the time the next.

Instead of sticking with the same rankings for top 4/mid 8/bottom 8, they could be recalculated after each game according to where they are on the ALT itself. The table would then feed back on itself and should be more accurate.

WARNING: That might be total bollocks

Can we have some more hilarious sketches please?

Oh feck

 

[Laphroaig]

Can we have an update please?

It takes a bit longer after losses!

 

[Plechazunga]

Update please twatheads

And can you make it a sticky please mods?

 

[MikeUpNorth]

It is done. Not looking too bad is it?

 

[alan_nis]

Now it's looking worse than the actual table.

 

[MikeUpNorth]

Updated.

 

[Plechazunga]

Can I suggest a refinement for next season?

The biggest inaccuracy as I see it is the initial rankings of teams. A side like West Ham can finish high up one season and easily be in the bottom 5 most of the time the next.

Instead of sticking with the same rankings for top 4/mid 8/bottom 8, they could be recalculated after each game according to where they are on the ALT itself. The table would then feed back on itself and should be more accurate.

WARNING: That might be total bollocks

Mike, what do you think of this, or something like it, for next season? It would take more work, but it is a flaw in the system. Liverpool got 2 points today for winning at Goodison, but Everton could end up in the bottom eight. If the rankings could be continually updated according to a similar set of expectations to the top 4's, wouldn't it be more accurate?

 

[MikeUpNorth]

Not sure that would work Plech, at least not until 10 or so games had been played. If a team starts with a few games that they're not expected to get anything from, and they lose as expected, then it's impossible to gauge their form. Those initial groupings are likely to be pretty accurate I'd say.

 

[pillory]

Mike, what do you think of this, or something like it, for next season? It would take more work, but it is a flaw in the system. Liverpool got 2 points today for winning at Goodison, but Everton could end up in the bottom eight. If the rankings could be continually updated according to a similar set of expectations to the top 4's, wouldn't it be more accurate?

I'm on it. (I suddenly remembered there's nothing I love more than geeking around with Excel, especially when I have a few bottles of wine and no deadlines.)

 

[MikeUpNorth]

Plech, I also think you can't adjust from one week to the next, or if you did you would have to go back through the previous weeks and amend the points as to where the teams are going to finish. Say we decide Everton are going to finish bottom 8 based on form, then we would have to adjust the previous results against them. That would work I guess, is that what you meant?

This is getting very nerdy. Good stuff.

 

[Plechazunga]

Not sure that would work Plech, at least not until 10 or so games had been played. If a team starts with a few games that they're not expected to get anything from, and they lose as expected, then it's impossible to gauge their form. Those initial groupings are likely to be pretty accurate I'd say.

Yeah as I said before like you'd have to start from last season's rankings.

Plech, I also think you can't adjust from one week to the next, or if you did you would have to go back through the previous weeks and amend the points as to where the teams are going to finish. Say we decide Everton are going to finish bottom 8 based on form, then we would have to adjust the previous results against them. That would work I guess, is that what you meant?

Er... yes you're right, it won't work. I think. Let's see what Pillory comes up with.

 

[SecondFig]

Plech, I also think you can't adjust from one week to the next, or if you did you would have to go back through the previous weeks and amend the points as to where the teams are going to finish. Say we decide Everton are going to finish bottom 8 based on form, then we would have to adjust the previous results against them. That would work I guess, is that what you meant?

This is getting very nerdy. Good stuff.

That's what you'd need, a nice excel chart that automatically recalculates points every week dependant upon the current league standings, thus it would become more accurate as the season goes on and teams get closer and closer to their final league places.

The only problem with that is it then simply replicates the actual league table, which as the season goes on becomes more and more accurate. The whole idea of the adjusted league table was that it placed a value upon matches at an early stage of the season.

 

[MikeUpNorth]

Exactly Mr Fig, exactly. I think Plech is talking bollocks myself. Not burst out laughing bollocks, but bollocks nonetheless.

 

[Plechazunga]

That's what you'd need, a nice excel chart that automatically recalculates points every week dependant upon the current league standings, thus it would become more accurate as the season goes on and teams get closer and closer to their final league places.

The only problem with that is it then simply replicates the actual league table, which as the season goes on becomes more and more accurate. The whole idea of the adjusted league table was that it placed a value upon matches at an early stage of the season.

That's right. But it's also meant to be predicated on pete's theory that you win the league by winning your home games, beating the bottom eight away and drawing with the top eight away. Instead, we're awarding points based on winning home games, beating some crap and some decent teams away, and drawing to some decent and some crap teams away.

What it needs is some sort of weighting. So if a club like Everton, say, which is expected to be top eight, is underperforming its expectations, the system would shave a bit off the amount you received for beating them away.

Mind you, there's a fine line between this and just saying can we unilaterally bump a point off the dippers please...

 

[SecondFig]

That's right. But it's also meant to be predicated on pete's theory that you win the league by winning your home games, beating the bottom eight away and drawing with the top eight away. Instead, we're awarding points based on winning home games, beating some crap and some decent teams away, and drawing to some decent and some crap teams away.

What it needs is some sort of weighting. So if a club like Everton, say, which is expected to be top eight, is underperforming its expectations, the system would shave a bit off the amount you received for beating them away.

Mind you, there's a fine line between this and just saying can we unilaterally bump a point off the dippers please...

Sounds good to me

 

[noodlehair]

Mind you, there's a fine line between this and just saying can we unilaterally bump a point off the dippers please...

Yeah, but in order to cross that line, you first have to cross the one that makes you realise the adjusted league table isn't actually real

 

[WeasteDevil]

This system is a bit spasticated isn't it?

 

[MikeUpNorth]

Yeah, but in order to cross that line, you first have to cross the one that makes you realise the adjusted league table isn't actually real

Well it works out mathematically that whoever finishes top of this wins the league.

 

[pillory]

Ok, here are my findings based on the league table as it stands right now.

Pete's system:
Chelsea 5
Man Utd 1
Arsenal -5
Liverpool -9

Rowem's system:
Chelsea 4.5
Man Utd -0.25
Arsenal -5
Liverpool -9.25

This is good news for us and bad news for Arsenal and Liverpool compared to the predetermined ranking stuff, so I suggest we stick with this method from now on.

 

[WeasteDevil]

It's stupid.