Certified Legendary Thread Race for the flag, in squiggly lines

[sverik25]

Just one question, it's probably already been asked, so apologies in advance...

Why is the the average score programmed as 85, when for the majority of recent years it has been in the low 90's? Would 90, for example, be better as a starting point?

 

[lusheslewis]

It will be interesting if we (Eagles) have a better than expected result against St. Kilda this week. If you move the Eagle a couple of points up and to the right you get this!

 

[Rob]

It will be interesting if we (Eagles) have a better than expected result against St. Kilda this week. If you move the Eagle a couple of points up and to the right you get this!

Way to ruin the squiggle.

Nice work.

 

[tazaa]

Way to ruin the squiggle.

Nice work.

I know it was a bit of fun but if that ever becomes the squiggles prediction than Final Sirens work would all have been for nothing.
Hawthorn would annihilate West Coast by 10 goals on any ground (no disrespect to Eagles).

 

[Pessimistic]

I think of the squiggle as a model representation of the vague pecking order you have in your mind

Except better of course

 

[AlibiMonday]

So you could argue that the final ladder is probably going to end up as a compromise between Squiggle and the current ladder.

By my interpretation, to compromise between the 2 would generally draw a more inaccurate result than the Squiggle alone. If anything - you should probably determine the discrepancy for any given team between the two, then move by 1-2 positions* in favour of the Squiggle's bias, rather pick a point in between.

* for this time of the year.

 

[clogged]

By my interpretation, to compromise between the 2 would generally draw a more inaccurate result than the Squiggle alone. If anything - you should probably determine the discrepancy for any given team between the two, then move by 1-2 positions* in favour of the Squiggle's bias, rather pick a point in between.

* for this time of the year.

Why? Both the Squiggle and the ladder are off by around 2 positions at round 7 on average.

Now, the actual ladder may have greater variance (last year teams like Collingwood and Gold Coast dropped seven spots while Geelong, West Coast and Melbourne ended up exactly where they were in round 7) against the Squiggle where teams predicted positions at the end of the season at round 7 are within 2-3 wins of their eventual position, but without ladders at the same stage of the season to compare, we don't know.

I know on Freo alone, the side went from predicted top of the ladder and premiers at round 17 to a prediction that had them barely making top four after the St Kilda loss. So even late in the season, with an accumulation of scores to weigh a side, and a handful remaining to determine the final outcome, one result can skew the prediction fairly easily.

 

[inbetween]

I know it was a bit of fun but if that ever becomes the squiggles prediction than Final Sirens work would all have been for nothing.
Hawthorn would annihilate West Coast by 10 goals on any ground (no disrespect to Eagles).

I would not say his work would have been for nothing.
Those predictions are made on a week to week, actually, a game to game basis. Theyre dynamically updated and so even if those were the predictions for even for a moment which is very possible, they're only the predictions for a brief moment in the AFL season and the following week it could have Hawthorn winning the flag yet again.

 

[Final Siren]

I think what it says is that the Squiggle isn't really a better predictor of end of season placings that the actual ladder in any given round aside from the very early ones, even though it attempts to control for things like easy or difficult draw. At round seven, for example, Squiggle is on average only 0.5 placings more accurate than the ladder in round 7. And neither are very accurate, they're both off by around 2 placings.

So you could argue that the final ladder is probably going to end up as a compromise between Squiggle and the current ladder. Which means despite Squiggle saying they'll finish second, Hawthorn are not an unlikely chance of missing top four, and while Squiggle thinks Port will finish in the eight "Round 7 ladder" says they'll miss finals by fair way.

Another way to look at it is to ask: How far ahead of the AFL ladder is the squiggle predictor? Pick a point on a chart and move sideways to see how many weeks it took for the AFL ladder to catch up.

For example, in 2014, it wasn't until after Round 7 that the AFL Ladder became as good a guide to finishing position as the squiggle was with its pre-season prediction. It took until after Round 12 for the AFL Ladder to reach where the squiggle was after Round 1... an 11-week lag. And it wasn't until Round 16 that the AFL ladder caught up to where the squiggle was after Round 3! But from Round 16 there was no real gap at all.

Running some numbers: over the last five years, the squiggle predictor is 2.4 weeks ahead on average. In the first 11 weeks of the season, it's 4.2 weeks ahead, and in the second half, it's 0.8 weeks ahead.

 

[Final Siren]

By my interpretation, to compromise between the 2 would generally draw a more inaccurate result than the Squiggle alone. If anything - you should probably determine the discrepancy for any given team between the two, then move by 1-2 positions* in favour of the Squiggle's bias, rather pick a point in between.

* for this time of the year.

Mmm, it would be interesting to look at. I'd guess that most prediction error is due to things like Richmond 2014, where both the squiggle and the AFL Ladder had them a lot lower than their eventual finishing place because they put on a burst of good form toward the end. In that case, there's indeed no compromise you could make between the ladder and the squiggle predictor that would give you a better result than one or the other alone, since the Tigers ended up placing higher than both suggested.

 

[tazaa]

I would not say his work would have been for nothing.
Those predictions are made on a week to week, actually, a game to game basis. Theyre dynamically updated and so even if those were the predictions for even for a moment which is very possible, they're only the predictions for a brief moment in the AFL season and the following week it could have Hawthorn winning the flag yet again.

With the top 4 all playing eachother early on its caused the squiggle to pump up teams and not rate others.
Obviously itll sort itself out after the byes but the fixture is to blame really.

 

[Truedogz]

Hi

I just joined up and really enjoy this thread and the squiggles.

Final Siren, I'll second what chunkychicken suggested. Extend back to say about 1980 the teams with their premiership cup symbols on the chart. I know it will make it a bit cluttered but it does better inform discussion on where the 'sweet spot' is as well how the different sides compare.

I'd also suggest adding a diagonal line representing the perfect balance between attack and defence.

As to what chunkychicken suggested the best all-round performance of the past 50 years was Hawthorn in 1988. There are four performances which are well above the rest of the pack and multiplying the attack/defence scores allows the greatest sides to be ranked:

Hawthorn 1988 = 86 x 79 = 6794
Geelong 2011 = 86 x 72 = 6192
Hawthorn 1989 = 89 x 69 = 6141
Essendon 1985 = 86 x 70 = 6020

I can remember the end of the great Melbourne era and have watched all the sides since. I think Geelong in 2011 was the best balanced team I have ever seen (and with a better attack than 2007-9) and rank that side and Hawthorn 1988-89 higher than Brisbane 2001-4 due to the quality of the opposition they had. Those two sides really were breathtaking to watch.

Hawthorn are in the box seat this year but their achilles heal, as over recent years, is their defence. If the Dockers and Swans can maintain their defensive strength and improve their attack they can take the prize.

Best Wishes

Truedogz

 

[Pessimistic]

One thing I noticed how tight and orderly the mass of squiggles look in recent years, yet around the eighties squiggles went all over the place

Is it a product of professionalism and or are games much more predictable these days

 

[Dipper]

One thing I noticed how tight and orderly the mass of squiggles look in recent years, yet around the eighties squiggles went all over the place

Is it a product of professionalism and or are games much more predictable these days

I'm guessing that higher scoring games provide a much greater range for both defence and attack. Fewer teams possibly makes it appear more polarised too.

 

[Randomizor]

The squiggle is going to love us this week

 

[SgtSchulz]

That is a fair ol' gulf that has formed now.

 

[Pessimistic]

If the grand final is sydney v freo, is it unprecedented for three teams to monopolise four grand finals ?

 

[treefingers]

If the grand final is sydney v freo, is it unprecedented for three teams to monopolise four grand finals ?

1983 - Hawthorn V Essendon
1984 - Hawthorn V Essendon
1985 - Hawthorn V Essendon
1986 - Hawthorn V Carlton
1987 - Hawthorn V Carlton

Three teams, Five grand finals

 

[SuperSub17]

If the grand final is sydney v freo, is it unprecedented for three teams to monopolise four grand finals ?

1979 Carl v Coll
1980 Rich v Coll
1981 Carl v Coll
1982 Carl v Rich

1975 Haw v NM
1976 Haw v NM
1977 Coll v NM
1978 Haw v NM


Then there was Rich-Carl in 72-73 and Rich-NM in 74, so from 1972 to 1987, only six teams played in those 16 grand finals

 

[Gydafud]

If the grand final is sydney v freo, is it unprecedented for three teams to monopolise four grand finals ?

Oh, I got some as well!

09 - Cats vs Saints
10 - Saints vs Piesx2
11 - Cats vs Pies

 

[Truedogz]

If the grand final is sydney v freo, is it unprecedented for three teams to monopolise four grand finals ?

1960 Melbourne Collingwood
1959 Melbourne Essendon
1958 Collingwood Melbourne
1957 Melbourne Essendon
1956 Melbourne Collingwood
1955 Melbourne Collingwood

Melbourne were runners up to Footscray in 54.

 

[jod23]

Final Siren, do you belong to Gryffindor, Ravenclaw, Hufflepuff or Slytherin?

I ask because I need to know which house to award 10 points too.

Bravo on this thread and your work. The squiggle treats my team very well so far.

 

[Roobs321]

GWS has now reached the point where comfortably edging a mid-table Squiggle side (and slightly outdoing tip) barely impresses Squiggle. Which is a compliment of sorts that they have arrived.

 

[The Swans Blog]

Squiggles has GWS making a semi. Decent start by them at 6-2 and a decent percentage.

 

[The Swans Blog]

If the grand final is sydney v freo, is it unprecedented for three teams to monopolise four grand finals ?

If it's Hawks vs Swans, then it's something like this over the last 10 years

Swans 2005
Swans 2006
Swans 2012
Swans 2014
Swans 2015
-------------
Hawks 2008
Hawks 2012
Hawks 2013
Hawks 2014
Hawks 2015