Certified Legendary Thread Race for the flag, in squiggly lines

[The_Steadier]

My statistical analysis is based on Franklins Per Team (FPT).
Sydney FPT= 1
Hawthorn FPT= 0

Swans win.

 

[Slawk]

My statistical analysis is based on Franklins Per Team (FPT).
Sydney FPT= 1
Hawthorn FPT= 0

Swans win.

Based on previous FPT's this gives Sydney a 66% chance of winning. Seems sound.

 

[MC_9]

Based on previous FPT's this gives Sydney a 66% chance of winning. Seems sound.

However based on FPT's in grand finals played between Sydney and Hawthorn, it is 0-1.

 

[Billy ray]

Algorithm details: For anyone interested in more nitty-gritty, here are some notable algorithms and their tips. Overall, 7 are tipping Sydney (by 6-18 pts) and 1 is tipping Hawthorn (by 9).

1. OFFDEF-75. My best Grand Final picker. No home advantage. Heavily weights recent games. 25-year GF record: 80%. 10-year GF record: 90%. Tip: Sydney by 10.
2. VENUE-81:13. Considers performances at the MCG. 25 years: 76%. 10 years: 80%. Sydney by 6.
3. OFFDEF-88:5. 5-pt home advantage to the nominal home team. Some more weight to recent games. 25 years: 80%. 10 years: 70%. Sydney by 11.
4. ISTATE-91:12: POWER:88. Heavily weights games against strong opposition. 25 years: 68%. 10 years: 80%. Sydney by 12.
5. PILGRIM-91:12. Penalizes teams based on distance travelled between games. 25 years: 72%. 10 years: 60%. Sydney by 8.
6. ISTATE-91:12. Core squiggle algorithm, using nominal home ground advantage. 25 years: 60%. 10 years: 50%. Sydney by 18.
7. ISTATEV2F-91:12. Core squiggle algorithm, using geographic home ground advantage. 25 years: 56%. 10 years: 60%. Hawthorn by 9.
8. HOMER. Tip the home team, i.e. the one that finished higher on the ladder. 25 years: 52%. 10 years: 50%. Sydney.

sooooooo, final siren, are you a betting man? if you were down to your last $10, and were going to have a bet, who and by what margin ?
the models are 7-1 in favour of Sydney and average margin is 11.

 

[clogged]

Our spread is also superior, the Swans have only had 5 players kick more than 20 goals for the season (and only 1 with more than 50), Hawthorn has had 7 (and 3 with more than 50 a piece) so that's another misconception - this isn't 2012 afterall where the Swans had 8 runners all with 20 or more goals.

Great minds. I will correct you though. Sydney had 9 players with 20+ goal tally going into the 2012 GF.

That's an amazing spread of goal kickers. This year it has nearly halved to five as you say.

Since 2012, Hawthorn have improved in ruck and defence. Sydney have added Franklin and Tippett, yet they have not become any more powerful and efficient in the forward 50. They are arguably weaker in the ruck.

 

[The King!]

Based on? Vibe?

This is a thread that looks at ways to predict winners by means of statistics. I've just proven that based on delivered output this is the closet. If you want to look at subjective measures, Hawthorn was clear favourites in 2012...a bit like the Swans this year, but on output the game was far closer (albeit still in the Hawks favour)



This isn't 2012. You don't have the spread you had in 2012, your forward set up I50 is completely different (as it is for Hawthorn). Hawthorn have a fair different game plan as is reflective in their statistics across all indicators. If you think that a game 2 years ago will provide an indicator as to the performance on Saturday you are a fool

Our top 6 goal kickers have missed a combined 29 games, hawthorns 8

Our spread is slightly deceptive given the forward line has rarely set up near its best, when our forwards are actually fit we score higher and more get involved, when they we tend to shut the game down more

I agree with your final sentence, thats a given, i would also say the last 24 week potentially tell bugger all too, either side could win this, no different to most grand finals, dont know about you but 2000 maybe 88 and 95 are about the only times i thought the result was easy as to pick

 

[Final Siren]

sooooooo, final siren, are you a betting man? if you were down to your last $10, and were going to have a bet, who and by what margin ?
the models are 7-1 in favour of Sydney and average margin is 11.

Not a betting guy, no. I have too much respect for the efficiency of markets for that. But I do look at odds sometimes as a good measurement of who everyone thinks is likely to win.

The thing I find fundamentally interesting about the squiggle is not that it's more accurate than the experts or the market, but rather that it's comparably accurate despite being very simple and ignoring everything except scores and venue.

 

[Hawkk]

The squiggle and Champion Data meet!

http://m.afl.com.au/news/2014-09-26/the-inquirer-grand-final-edition

Squiggle predicts Swans by 10
Champion Data predicts Swans by 4
Parko predicts Swans 18.5 / 17.5 over the Hawks

As long as the Hawks are in the game with 5 minutes left and it is a memorable GF I'll be happy

Sorry to get all subjective on you all but if the game is a coin flip late in the game I'll back us in. In much the same way that the Swans would comfortably back them in as well.

Here's hoping that the statistics align and we have another classic GF!

 

[The King!]

The squiggle and Champion Data meet!

http://m.afl.com.au/news/2014-09-26/the-inquirer-grand-final-edition

Squiggle predicts Swans by 10
Champion Data predicts Swans by 4
Parko predicts Swans 18.5 / 17.5 over the Hawks

As long as the Hawks are in the game with 5 minutes left and it is a memorable GF I'll be happy

Sorry to get all subjective on you all but if the game is a coin flip late in the game I'll back us in. In much the same way that the Swans would comfortably back them in as well.

Here's hoping that the statistics align and we have another classic GF!

i know you dislike my side and thats for another thread or whatever but i think your footy posts are very impressive.

i will back us in a tight one, but yeah i think it could be a coin flip, might be a heartbreaking loss for whoever is on the wrong flip side

 

[Final Siren]

The squiggle and Champion Data meet!

http://m.afl.com.au/news/2014-09-26/the-inquirer-grand-final-edition

Squiggle predicts Swans by 10
Champion Data predicts Swans by 4
Parko predicts Swans 18.5 / 17.5 over the Hawks

As long as the Hawks are in the game with 5 minutes left and it is a memorable GF I'll be happy

Sorry to get all subjective on you all but if the game is a coin flip late in the game I'll back us in. In much the same way that the Swans would comfortably back them in as well.

Here's hoping that the statistics align and we have another classic GF!

Huh, a 4-pt tip, that's one dodgy free kick. That's basically a 50/50 call!

Interesting that Champion Data generate their match tip by "comparing the selected players' individual ratings." That's essentially the opposite of the squiggle, which completely ignores individual players to rate overall team performance. And yet the tips are within a goal of each other.

 

[Hawkk]

i know you dislike my side and thats for another thread or whatever but i think your footy posts are very impressive.

i will back us in a tight one, but yeah i think it could be a coin flip, might be a heartbreaking loss for whoever is on the wrong flip side

I dislike the Swans coz the Swans are the one team that caused me the most heartbreak (potentially x2 tomorrow) in my time supporting football. Same reason why Geelong fans hate Hawthorn and Collingwood hate Brisbane Lions!

Either way good luck, just not too much if you get what I mean!

 

[deflated]

Great minds. I will correct you though. Sydney had 9 players with 20+ goal tally going into the 2012 GF.

That's an amazing spread of goal kickers. This year it has nearly halved to five as you say.

Since 2012, Hawthorn have improved in ruck and defence. Sydney have added Franklin and Tippett, yet they have not become any more powerful and efficient in the forward 50. They are arguably weaker in the ruck.

I suspect Sydney has taken advantage of the influx of forward talent in an unexpected manner.

The 2014 Swans are a far better team defensively than the 2012 edition. They have managed to keep a similar level of scoring while requiring fewer players to kick multiple goals. Adding star forwards wasn't seen as a chance to score more points but as an opportunity to shift the midfield to more of a defensive mindset and out of the F50.

Anecdotally it seems like JPK almost never drifts forward to pop up as a F50 target now and that was such a distinctive part of his game in 2012. The '4 talls' setup looks like it is focused on getting a 1-on-1 with whichever of them has the best matchup where the 2012 setup seemed to involve the mids leading into the F50 with kicks to open space in front of them. You work with what you've got.

 

[Pessimistic]

The squiggle and Champion Data meet!

http://m.afl.com.au/news/2014-09-26/the-inquirer-grand-final-edition

Squiggle predicts Swans by 10
Champion Data predicts Swans by 4
Parko predicts Swans 18.5 / 17.5 over the Hawks

As long as the Hawks are in the game with 5 minutes left and it is a memorable GF I'll be happy

Sorry to get all subjective on you all but if the game is a coin flip late in the game I'll back us in. In much the same way that the Swans would comfortably back them in as well.

Here's hoping that the statistics align and we have another classic GF!

Swans current squiggle is closest to where freo were last year

 

[Pessimistic]

I suspect Sydney has taken advantage of the influx of forward talent in an unexpected manner.

The 2014 Swans are a far better team defensively than the 2012 edition. They have managed to keep a similar level of scoring while requiring fewer players to kick multiple goals. Adding star forwards wasn't seen as a chance to score more points but as an opportunity to shift the midfield to more of a defensive mindset and out of the F50.

Anecdotally it seems like JPK almost never drifts forward to pop up as a F50 target now and that was such a distinctive part of his game in 2012. The '4 talls' setup looks like it is focused on getting a 1-on-1 with whichever of them has the best matchup where the 2012 setup seemed to involve the mids leading into the F50 with kicks to open space in front of them. You work with what you've got.

Is billed as the best defence syd v the best offence hawks

However the current listed teams have a better offensive and defensive record respectively. The stats are from the season as a whole

 

[Procrastinator35]

Swans current squiggle is closest to where freo were last year

Yeah....But Freo still had a way tougher draw than them last year!....The squiggle does not account for AFL Draw inequities, which is its downfall!

 

[The Swans Blog]

The squiggle and Champion Data meet!

http://m.afl.com.au/news/2014-09-26/the-inquirer-grand-final-edition

Squiggle predicts Swans by 10
Champion Data predicts Swans by 4
Parko predicts Swans 18.5 / 17.5 over the Hawks

As long as the Hawks are in the game with 5 minutes left and it is a memorable GF I'll be happy

Sorry to get all subjective on you all but if the game is a coin flip late in the game I'll back us in. In much the same way that the Swans would comfortably back them in as well.

Here's hoping that the statistics align and we have another classic GF!

It'll be a bit like the '11 GF where it was down the wire until 3QTR time and then Geelong blew it away. I think the same will happen and one team will charge ahead. It might be in the 4th, or it might be in the 2nd. Either way, one team is going to win by more than a goal. Which... well, I don't know.

 

[ScrewInStuds]

I suspect Sydney has taken advantage of the influx of forward talent in an unexpected manner.

The 2014 Swans are a far better team defensively than the 2012 edition. They have managed to keep a similar level of scoring while requiring fewer players to kick multiple goals. Adding star forwards wasn't seen as a chance to score more points but as an opportunity to shift the midfield to more of a defensive mindset and out of the F50.

Anecdotally it seems like JPK almost never drifts forward to pop up as a F50 target now and that was such a distinctive part of his game in 2012. The '4 talls' setup looks like it is focused on getting a 1-on-1 with whichever of them has the best matchup where the 2012 setup seemed to involve the mids leading into the F50 with kicks to open space in front of them. You work with what you've got.

I suspect that this the main reason that we look so fresh at the pointy end of the season this year - slingshot footy in comparison seems to be very hard on the body. Great analysis.

 

[Final Siren]

Yeah....But Freo still had a way tougher draw than them last year!....The squiggle does not account for AFL Draw inequities, which is its downfall!

Actually it does.

 

[Procrastinator35]

Actually it does.

Right, so if Sydney played Freo & The Cats away from home, as their top 4 finishing spot the 2 seasons prior dictated that they should have, then it's a fair estimate to say that they would not have finished first, much less second this season!.....Pray tell me how such a predicament as this is 'factored in'!!!???

Surely, such an anomalous contingency on the part of the AFL 're-writing' the prescription pad cannot be apprehended/digested by your Squiggle!

 

[Hawkk]

It'll be a bit like the '11 GF where it was down the wire until 3QTR time and then Geelong blew it away. I think the same will happen and one team will charge ahead. It might be in the 4th, or it might be in the 2nd. Either way, one team is going to win by more than a goal. Which... well, I don't know.

...and? Aside from 2007 all GF have been in the balance at 3/4 time and that dates back to 2003 so its not a recent phenomena at all. Based on produced performance the game is basically a coin flip (this is the unique aspect of this match up) but that's not to say one team won't get a leg up or play a vastly superior game in the GF and win by 8-10 goals (and they would be a deserved premier in that instance). Like the squiggle or champion data all predications are a reflection of past form and not a direct indicator of future performance.

The beauty about Australian Football is that quintessentially the premier is decided on the output of one game at a neutralised venue (14000 tickets each). Wayne Carey wrote what I believe to be the only half decent analysis for the Grand Final when he claimed that in any given year both teams have the opportunity to win the flag with an adequate number or I50s and scoring shots but its the side that converts these opportunities that wins the premiership. The reference he gave included:

2008 - Hawks over Cats (scoring shots 34 to 25 Cats way)
2009 - Cats over Saints (scoring shots 23 to 18 Saints way)
2010 - DRAW (scoring shots 23 to 18 Collingwood way)
2012 - Swans over Hawks (scoring shots 26 to 21 Hawks way)
2013 - Hawks over Dockers (scoring shots 22 apiece)

I just finished watching the 2012 GF (a game I prefer over the 2013 GF even to this day) and it dawned on me that missing shots on goal has both a sapping impact on the missing team as well as a confidence build for the opposition.

You don't make a GF unless you're capable of winning a game at any venue if your opponent continues to miss chances. Today won't be any different. Whoever maximises their chances will win the flag, period

As an aside its a bit of a slant on Australian journalism that Carey put the others to shame but it is what it is...

 

[spannerSM]

Pray tell me how such a predicament as this is 'factored in'!!!???

The squiggle doesn't just look at the result, it compares the actual scoreline to the predicted scoreline. When a higher-performing side plays a lower-performing side, it's not enough simply for the higher-performing side to win, to maintain their squiggle position they have to beat the line, basically.

 

[WWSD]

My statistical analysis is based on Spangher's Per Team (SPT).

Hawthorn SPT = 1
Sydney SPT = 0

Hawthorn win.

EFA.

 

[Sydney Bloods]

EFA.

but we have a cunningham.

 

[Hawkk]

Not a betting guy, no. I have too much respect for the efficiency of markets for that. But I do look at odds sometimes as a good measurement of who everyone thinks is likely to win.

The thing I find fundamentally interesting about the squiggle is not that it's more accurate than the experts or the market, but rather that it's comparably accurate despite being very simple and ignoring everything except scores and venue.

Tend to agree...

I posted this on the favourites thread but it makes for an interesting comparison

2014 - Swans 1.58 v Hawks 2.45 (Sportsbet, odds vary per bookmaker)

The last 5 GF's

http://www.oddsportal.com/aussie-rules/australia/afl-2014/results/

2013 - Hawks 1.61 v Freo 2.33 (Hawks by 15)
2012 - Hawks 1.55 v Swans 2.55 (Swans by 10)
2011 - Coll 1.98 v Geel 1.78 (Geel by 39 points)
2010 R - StK 2.36 v Coll 1.57 (Coll by 56 points)
2010 - StK 2.95 v Coll 1.39 (DRAW)
2009 - StK 2.31 v Geel 1.58 (Geel by 12 points)

I can also add proudly (cause I made something on it)

2008 - Geel 1.42 v Haw 2.80 (Haw by 26 points)

We've had an amazing run of GF's but a clear favourite in all GF's. Its somewhat surprising that the only close betting market we've had in recent years was the blow out in 2011 (in the last quarter!). I don't endorse gambling but I have to agree that it gives a pretty solid indication where the sentiment for the day lies.

I think the market is heavily skewed by the media and given the importance of the media developing a narrative for the day and the week the odds can get swept up. As more media covers the game and pundits (who are no different to you and me) try to accentuate strengths and weaknesses, the odds bounce around like a yo-yo. If you're in the media no one in the immediate family should be allowed to bet as opinion heavily slants the market.

A classic example is 2012 where the Age posted an article raising question with respect to a future Hawthorn dynasty BEFORE the GF. The reality was that we weren't much better than the Swans and played catch up for the most part to stay in the top 2. FFS the Swans spent more time on top that season than anyone else! We went in as hot favourites (even more favoured than this year) and were exposed.

The media needed a narrative and we were the glamour side at the time (probably skewed by our offence and ridiculous percentage - 155% from memory?). We were the definition of flat track bullies and the Swans won a deserved premiership (they smoked us in the 2nd). As it is, I'm kinda glad we went through that experience as I don't believe we would have won 2013 and made the necessary changes in 2014 if the Swans didn't highlight some of our glaring deficiencies (one dimensional forward line, lack of midfield spread and poor KPD posts)

On a that very note, because so much of the narrative this week leads back to 2012, there is some symmetry between 2012 and 2014 in the betting market... not a lot but some. Im not saying we'll win but the odds are very eery! I don't believe lightening strikes twice so write it off as something that is unusual but completely meaningless to the result at hand

 

[Final Siren]

The squiggle doesn't just look at the result, it compares the actual scoreline to the predicted scoreline. When a higher-performing side plays a lower-performing side, it's not enough simply for the higher-performing side to win, to maintain their squiggle position they have to beat the line, basically.

Right. So to hold your current chart position against a weak team, you might need to win by five goals, whereas against a strong team, a 2-goal loss might be good enough.

The other thing is that interstate home advantage is worth 12 points. So the Swans winning by 2 goals at home is seen as an equivalent performance to them losing by 2 goals away.

It doesn't matter if you have a hard draw or an easy one, play all your matches at home or away: the squiggle will account for this.