Certified Legendary Thread Squiggle 2017

Final Siren

Mr Squiggle
Aug 18, 2009
3,867
16,293
AFL Club
Richmond
Continued in thread for 2018/2019: https://www.bigfooty.com/forum/thre...reciation-thread-and-other-analytics.1187002/

======================================

Chief asked me to start a new thread this year so this is it.

Thanks for joining Squiggle 2017! This is the continuation of a journey that started here, with a chart of squiggly lines representing the race for the flag.

Here's the end result from the last few years:








Ahhh, 2016, you so crazy.

Squiggles are based on a simple model that you can read more about under INFO here. It looks only at scores and home ground advantage. Despite this, it's a fairly solid tipper, averaging 73% over the last 5 years, which goes to show how far you can go with only a little information. And it's a good excuse to talk about footy in an analytical way but without getting too technical.

Plus they're super pretty.

Here's the starting point for 2017! Positions are calculated from 2016 performance only: They don't take into consideration list changes, injuries, the fact that 90% of GWS is made up of first round draft picks just coming into their prime, and so on.


And here's the result of a run through the squiggle ladder predictor:


This has been a tumultuous off-season, with Essendon regaining half a side, North axing one, and Ross Lyon learning to care about winning again. So there may be more teams than usual moving far up and down the ladder compared to usual. But going on 2016 form alone, this is how they line up.

I want to say up-front that I'm pretty sure I'm going to spend this year charting how awesome GWS are, especially compared to my own team. So each weekly update will be like swallowing needles. But I'm prepared to do it.

Live squiggle! Play through past years, drag teams to where you think they belong and generate new season forecasts, and play what-if with Squiggle Sliding Doors!
 
Last edited by a moderator:

(Log in to remove this ad.)

(Log in to remove this ad.)

Final Siren

Mr Squiggle
Aug 18, 2009
3,867
16,293
AFL Club
Richmond
Final Siren shouldn't the dogs (and the rest of the teams) starting position be where they finished in 2016 (i.e position after the GF)
Teams usually move toward the middle (50/50) between seasons, because I drop one year's worth of results. The Dogs' 2016 finishing position is comprised of their 2015 + 2016 games, but their starting position is based on 2016 only. By the end of the year, it will be based on their 2016 + 2017 games.
 

Senor Ed

Club Legend
Oct 2, 2007
1,597
3,941
Melbourne
AFL Club
Western Bulldogs
Other Teams
Doggies
Teams usually move toward the middle (50/50) between seasons, because I drop one year's worth of results. The Dogs' 2016 finishing position is comprised of their 2015 + 2016 games, but their starting position is based on 2016 only. By the end of the year, it will be based on their 2016 + 2017 games.
Cheers, thanks for clearing that up.
 

Clems Knee

Norm Smith Medallist
Aug 15, 2009
7,857
14,745
Perth
AFL Club
Fremantle
Thanks for the squiggle. I hope the teams start moving their positions early in the season (especially my team).

I tried to drag all teams to 60 for attack and 60 for defence, to see how much the fixture benefitted teams, but the predictor wouldn't recalculate the season. Any idea of how to do this?
 

igon value

Club Legend
Oct 10, 2016
2,133
2,224
AFL Club
Fremantle
Final Siren - do you have a screen shot of the starting point of last year? I want to see how much things drifted from before round 1 to final result. IIRC not much changed (eg no one from bottom four surprised on the upside, or vice versa with the top four).
 

Final Siren

Mr Squiggle
Aug 18, 2009
3,867
16,293
AFL Club
Richmond
Thanks for the squiggle. I hope the teams start moving their positions early in the season (especially my team).

I tried to drag all teams to 60 for attack and 60 for defence, to see how much the fixture benefitted teams, but the predictor wouldn't recalculate the season. Any idea of how to do this?
Sorry, dragging & recalculating should work now.

With fixture bias, the major source (for squiggle purposes, at least) is which interstate teams you get at home vs away. Ideally you want your away interstate games to be matches you would have won or lost at home anyway, while your home interstate games are against evenly-matched opponents where the venue makes a critical difference.

That's not something you can check by dragging teams around, though. I think it's a very even fixture based on number of home/away interstate games alone. I get everyone on 11 wins when I put them at 50/50 and run through the season.
 

Final Siren

Mr Squiggle
Aug 18, 2009
3,867
16,293
AFL Club
Richmond
Final Siren - do you have a screen shot of the starting point of last year? I want to see how much things drifted from before round 1 to final result. IIRC not much changed (eg no one from bottom four surprised on the upside, or vice versa with the top four).
GWS moved a long way last year:


You can see this by hovering over their flag on the 2016 live squiggle. And you can see the starting positions for that year by using the player controls at the top.
 

LittleG

Norm Smith Medallist
Nov 18, 2015
5,177
5,638
AFL Club
Western Bulldogs
Squiggle predicting Longmire beating Beverage in all three meetings!
When was the last time Longmire beat beverage?
 

Pessimistic

TheBrownDog
Sep 13, 2000
73,933
32,361
Melbourne cricket ground. Australia
AFL Club
Hawthorn
Other Teams
Horks
Greater Western Sydney 4.50
Western Bulldogs 5.50
Sydney 7.00
Adelaide 9.00
West Coast 11.00
Geelong 11.00
Hawthorn 13.00
Fremantle 18.00
Melbourne 23.00
Essendon 23.00
St Kilda26.00
Collingwood 34.00
Port Adelaide 34.00
Gold Coast 41.00
Richmond 51.00
North Melbourne 67.00
Carlton 101.00
Brisbane 501.00

Punters not that far off the squiggle - a fair gap between 7 and 8 and the rest

Now you can believe the seasons nearly here
 
Last edited:

Top Bottom