Certified Legendary Thread Race for the flag, in squiggly lines

Final Siren

Mr Squiggle
Aug 18, 2009
4,229
17,495
AFL Club
Richmond
Here's a weird question Final Siren.

What are the likelihoods of an upset in the season/finals if the streams have or have not crossed?
That is a weird one. I don't have a way of generating an answer, other than peering at squiggles. I can see that in the 2014 finals, we've had two upsets, and in one of them the squiggles hadn't crossed (Freo vs Port) and in the other, they had (Geelong vs North).

I wouldn't imagine there'd be much in whether teams had crossed paths during the season or not.
 

Final Siren

Mr Squiggle
Aug 18, 2009
4,229
17,495
AFL Club
Richmond
Possible to post an updated version of this 50/50 Squiggle Final Siren?
Sure! By now there's not much difference between that chart and the main squiggle, since the weighting of older games diminishes by 9% per round. So while at the start of the year, a team's position is 100% based on 2013 results, after Round 1 it's 91% based on 2013 results, then after Round 2 it's 83% (0.91 x 0.91 = 0.8281), then 75%... and so on. After 22 games, the previous year's results are only 12.6%.

For teams that have finished their seasons, this will be their 2015 starting position.

0EcsxRU.png
 

Final Siren

Mr Squiggle
Aug 18, 2009
4,229
17,495
AFL Club
Richmond
i know the squiggle works on attack and defence but is it possible to have a reference point? like a mark that shows roughly a median of what is the winning formula in any given year. I dont know if its possible at all but in seasons gone by mabe kicking 120 while conceding 90 was the comp average and the eventual premier bucked that trend? Currently we say well the 2 sides are similar in attack but one is ahead in defence but what if this year the trend is more towards attack (over the whole comp)
I don't see any real pattern there... for example, 2000 was a super-attacking year, and the flag was won by Essendon, the most attacking team. But 2012 was pretty attacking, too, and the flag was won by Sydney, the most defensive team.

What is notable, though, is how defensive this year is: it's the first time since 1968 that the average DEFENCE is higher than average ATTACK. That was also the last time that average scores were as low as they are now (82 pts in 1968, 86 pts now). In most of the intervening years, average ATTACK was a lot higher, with teams scoring over 100 points on average. In the late 90s the gap began to close, with average scores dipping below 100, and there were a few years this century where average ATTACK and DEFENCE got close. But it hasn't ticked over until now.
 

KillaKanga

Norm Smith Medallist
Feb 9, 2011
5,990
6,462
If I'm online I'm at work!
AFL Club
North Melbourne
Other Teams
Boston CelticsRedSoxBruins
Really? i've been called out for declaring my team won't win games. now for declaring they will.
are you saying you've never declared on bigfooty your team would win a game before it was actually played?

I never said you didn't have a shot, I've said "it will be be a Tough game, but we will come away with the bickies."
Please forgive me for great crime of believing the team that finished first will beat the team that finished 6th. :rolleyes:
I've never declared my team WILL win a game. I've used should, could, 'I'll be pi$$ed if we lose' but never WILL. How can anyone be so sure of an outcome they have no impact on. Even the Squiggle can't make those declarations.

Also may be good to look at the context of my post and who/why I was quoting ;)
 
Last edited:

Mayes2Gray

Club Legend
May 18, 2014
1,267
1,185
AFL Club
Port Adelaide
Sure! By now there's not much difference between that chart and the main squiggle, since the weighting of older games diminishes by 9% per round. So while at the start of the year, a team's position is 100% based on 2013 results, after Round 1 it's 91% based on 2013 results, then after Round 2 it's 83% (0.91 x 0.91 = 0.8281), then 75%... and so on. After 22 games, the previous year's results are only 12.6%.

For teams that have finished their seasons, this will be their 2015 starting position.

0EcsxRU.png
the collingwood and west coast lines surprise me for some reason. I'm sure there is a logical answer, but supriding nonetheless
 

Pessimistic

Cancelled
30k Posts 10k Posts HBF's Milk Crate - 70k Posts TheBrownDog
Sep 13, 2000
86,852
42,951
Melbourne cricket ground. Australia
AFL Club
Hawthorn
Other Teams
Horks
upload_2014-9-18_20-53-42.png

Put some more work into my grand final vectors. the dot represents the loser and the arrow the winner of the grand final for that year. This contains all the vectors which point NW to NE. Nw suggests an attacking rated team beating a defensive rated team. when the vector points NE, it represents a stronger team playing a middle of the road side. here came some of the bog blowouts, Essendon 1985, Hawks 1988, Collingwood 1990 Essendon 2000 Geelong 2007. seems to happen once per decade. Maybe one due in the near future but the current squiggles suggest not this year.

Due reference to the squiggle father by the way
 

Attachments

  • upload_2014-9-18_20-41-37.png
    upload_2014-9-18_20-41-37.png
    66.5 KB · Views: 8

JZTIME!

All Australian
Jul 21, 2011
883
617
1999
AFL Club
North Melbourne
I don't see any real pattern there... for example, 2000 was a super-attacking year, and the flag was won by Essendon, the most attacking team. But 2012 was pretty attacking, too, and the flag was won by Sydney, the most defensive team.

What is notable, though, is how defensive this year is: it's the first time since 1968 that the average DEFENCE is higher than average ATTACK. That was also the last time that average scores were as low as they are now (82 pts in 1968, 86 pts now). In most of the intervening years, average ATTACK was a lot higher, with teams scoring over 100 points on average. In the late 90s the gap began to close, with average scores dipping below 100, and there were a few years this century where average ATTACK and DEFENCE got close. But it hasn't ticked over until now.
View attachment 80817
Put some more work into my grand final vectors. the dot represents the loser and the arrow the winner of the grand final for that year. This contains all the vectors which point NW to NE. Nw suggests an attacking rated team beating a defensive rated team. when the vector points NE, it represents a stronger team playing a middle of the road side. here came some of the bog blowouts, Essendon 1985, Hawks 1988, Collingwood 1990 Essendon 2000 Geelong 2007. seems to happen once per decade. Maybe one due in the near future but the current squiggles suggest not this year.

Due reference to the squiggle father by the way
love it, would be interesting to see with just 2 different coloured dots
 

Pessimistic

Cancelled
30k Posts 10k Posts HBF's Milk Crate - 70k Posts TheBrownDog
Sep 13, 2000
86,852
42,951
Melbourne cricket ground. Australia
AFL Club
Hawthorn
Other Teams
Horks
upload_2014-9-18_21-11-3.png

I put some more work into my grand final vectors. the dot is the loser and the arrow is the grand final winner. These are the ones where a defensive team beats an attacking rated team. About one third of the last 40 grand finals ended this way.
There is one outlier from those pointing from NE to SE. The 2008 grand final, but if you look at 2008, Hawks were ahead of the pack too and in premiership territory. So the grand finals where the defensive rated teams triumphed are:

2012 Sydney def Hawthorn
2008 Hawthorn def Geelong
2005 Sydney def WCE
1998 Adelaide def North
1997 Adelaide def St Kilda
1995 Carlton def Geelong (not sure that was an upset)
1994 WCE def Geelong
1992 WCE def Geelong
1989 Hawthorn def Geelong (though both teams were in all time best squiggle territory)
1987 Carlton def Hawthorn
1984 Essendon def Hawthorn
1981 Carlton def Collingwood

Interestingly none of these happened in the seventies from five years and only once from five years in this decade. Geelong and Hawthorn feature a lot on the receiving end, and carlton, Hawthorn, WCE, Adelaide, Sydney the main upsetters
 

Final Siren

Mr Squiggle
Aug 18, 2009
4,229
17,495
AFL Club
Richmond
View attachment 80817
Put some more work into my grand final vectors. the dot represents the loser and the arrow the winner of the grand final for that year. This contains all the vectors which point NW to NE. Nw suggests an attacking rated team beating a defensive rated team. when the vector points NE, it represents a stronger team playing a middle of the road side. here came some of the bog blowouts, Essendon 1985, Hawks 1988, Collingwood 1990 Essendon 2000 Geelong 2007. seems to happen once per decade. Maybe one due in the near future but the current squiggles suggest not this year.

Due reference to the squiggle father by the way
That's pretty awesome! Maybe you could colour the lines by direction, so it would be obvious at a glance roughly how many are going NE vs NW vs SE vs SW?
 

JZTIME!

All Australian
Jul 21, 2011
883
617
1999
AFL Club
North Melbourne
just dot the winners and losers in different colours, a correlation should appear and we can see how it relates to positions on the squiggle
 

Final Siren

Mr Squiggle
Aug 18, 2009
4,229
17,495
AFL Club
Richmond
just dot the winners and losers in different colours, a correlation should appear and we can see how it relates to positions on the squiggle
It is useful to see the pairs, though. For example, in 2011 Collingwood lost from a very good squiggle position, but to a Geelong team that charted even better.
 

Final Siren

Mr Squiggle
Aug 18, 2009
4,229
17,495
AFL Club
Richmond
the collingwood and west coast lines surprise me for some reason. I'm sure there is a logical answer, but supriding nonetheless
They both moved a lot in the last month, especially Collingwood, who had an abysmal finish. Their final position is not a summary of their performance over the course of the year but rather a measure of their form at the end.
 

Pessimistic

Cancelled
30k Posts 10k Posts HBF's Milk Crate - 70k Posts TheBrownDog
Sep 13, 2000
86,852
42,951
Melbourne cricket ground. Australia
AFL Club
Hawthorn
Other Teams
Horks
It'd be interesting to see what you'd get if you added all the vectors together.

Not sure about that, but this is the grouping

(If Dan26 is around, how does this tie to your premierships in order of awesomenesss)

upload_2014-9-18_22-14-56.png

Im also sensing that if the grand final combatants are either more defensive or more attacking, tose arrows aligned nw to se can go either way about 50 50.

But if one team is superior in defance and attack, particularly close to the SW NE axis, that team will win 90% of the time, maybe more

There are cases pointing almost due north, where both teams are defensively similar but one has superior attack. This team always wins


Also if this weeks games were grand finals, hawks def Port would be the statistical favourite, and syd v north could go either way
 

Attachments

  • upload_2014-9-18_22-9-2.png
    upload_2014-9-18_22-9-2.png
    30.8 KB · Views: 9
Last edited:

Final Siren

Mr Squiggle
Aug 18, 2009
4,229
17,495
AFL Club
Richmond
Do we have a top 8 squiggle? How much are the results inflated by blowouts vs pretenders?
In theory, not much, since the quality of opposition is already taken into account. But there are definitely cases where teams get a very big boost from holding a weak team to a very low score.

So I'll take a look. I'm not usually a fan of stats "vs the top 8" (or 4), since they severely reduce the data set, which increases potential error, and it's so arbitrary - you weight results against a team that won 12 games at 100% but results against a team that won 11 games at 0%!

But there probably are some interesting differences to be found there. I suppose North would look a lot different.
 
In theory, not much, since the quality of opposition is already taken into account. But there are definitely cases where teams get a very big boost from holding a weak team to a very low score.

So I'll take a look. I'm not usually a fan of stats "vs the top 8" (or 4), since they severely reduce the data set, which increases potential error, and it's so arbitrary - you weight results against a team that won 12 games at 100% but results against a team that won 11 games at 0%!

But there probably are some interesting differences to be found there. I suppose North would look a lot different.

Yeah, I am just curious if there is a significantly different pattern.
 

Dipper

Norm Smith Medallist
Oct 28, 2000
7,984
3,364
London,England
AFL Club
Hawthorn
Other Teams
Allies FCC
Not sure about that, but this is the grouping

View attachment 80841

Surely the angle not the direction of the arrows is key.(for anything running NW-SE)

If an arrow goes NE, it means a side superior in both attack and defence won.

If it's N, then same defence but better attack.

So these would be completely expected, any arrows going SW mean that a side inferior in both defence and attack has lost, which we obviously don't expect and that's why there is only one example.


The area of interest is arrows going NW and SE, you seem to be looking for a very simple answer, does the better attacking or defensive side usually win.

But the key is the relative difference between the two side's attack/ defence ratio.

So when you have an arrow going from the NNW quadrant to the SSE quadrant that shows that the side placed there is relatively stronger in its attack than the opposition's defence, and so the odds would be on them winning.

Similarly if the sides are placed in the NWW and SEE in relation to one another then the defensive side has a better defence relative to the opposition's attack.

So you wouldn't expect too green many arrows in the SSE quadrant and you wouldn't expect many red arrows in the NWW quadrant.(especially nearer the extremes)

And what you see is that the results fit in with expectations to a large degree.

View attachment 80817
when the vector points NE, it represents a stronger team playing a middle of the road side.

It doesn't necessarily mean that, one side could be the smallest possible increment better than another in both attack and defence and they will be to the NE of them but in reality they would be very close in ability.(the length of the line determines the relative strengths when going NE/ SW)

Contrast this with a side that is the slightest increment worse defensively than another but is way better in attack.....that arrow will go to the NW (but in reality it will be as near as dammit N) yet the difference in strength between those 2 sides will be much bigger than the previous example with a NE direction.


It seems to me that the further any side is away from the bottom left hand corner the better they are, it's that simple,
 

Dipper

Norm Smith Medallist
Oct 28, 2000
7,984
3,364
London,England
AFL Club
Hawthorn
Other Teams
Allies FCC
It is useful to see the pairs, though. For example, in 2011 Collingwood lost from a very good squiggle position, but to a Geelong team that charted even better.

Did Geelong actually chart better than Collingwood that year though?

Collingwood were out there beyond even WCE1994 for defence and yet were still around the median premiership winning range for attack, and are further from the bottom left pivot than Geelong, which surely says that their attack/defence ratio was superior.....or am I getting things drastically wrong.......I'm very hesitant to disagree with the genius behind the squiggle.
 
Back