Certified Legendary Thread Race for the flag, in squiggly lines

[Mad_Hatter]

Little question Final Siren , and I don't really need a full on explanation or detailed statistical breakdown but moreso just your opinion, but how would Geelong look on the squiggle and flagpole if we didn't drop those games to lesser teams?
I feel considering the Cats record vs top 8 sides that it should be higher, no doubt brought down by those crappy loses.
Would our position improve all that much?

 

[The Swans Blog]

Rare but not unprecedented. 80+ Defence at any point in the last 20 years:

• Fremantle 2014 Rounds 16-17
• Collingwood 2011 Rounds 20+
• St Kilda 2010 Round 6
• St Kilda 2009 In and out across the year
• Adelaide 2009 Round 15 (one game only, after holding Fremantle to 13 points)
• Adelaide 2005 Semi-final & prelim

Certainly not unprecedented in defence, but certainly in the average age and average experience category

 

[Rise_up_Stiffy]

Certainly not unprecedented in defence, but certainly in the average age and average experience category

Swans have been amazing all year, looks like the majority of the media will finally be jumping on board this week. 3 Horse race I reckon, Swans, Crows & Hawks (in that order)

 

[Tippett2Goal]

North play GWS & Sydney. According to the ladder predictor they will win 1 of 2, but predictor is also saying that Swans and Giants both win both of theirs.

Am I missing something?

Tips are probabilities, so say North are currently on 12.2 wins (rounded down to 12), and are 70/30 chances vs both Swans and GWS, the predictor will give them .3 of a win for each of the upcoming games moving them to 12.8 (rounded up to 13). With each of Sydney and GWS getting .7 of a win added to their scores. (and similarly for their other game, which will round up to 2 wins)

 

[Rotayjay]

Round 21, 2016

Animated!

This week it started to resemble 2013, when the Grand Final was between the league's best scorer (Hawthorn) and its best stopper (Fremantle). Sydney are actually about as defensive as teams ever get, with only a handful of teams treading this territory over the last 20 years. And it's been a great three weeks for them in terms of firepower, too, scoring 137 against Fremantle, 100 against Port, and 146 against St Kilda.

It was a good week for everyone else in the top 5 as Greater Western Sydney won this week's episode of WHO WANTS TO DROP A CRUICIAL GAME? (with a gallant effort from Geelong). This meant the Tower of Power, which last week had an even 5-way split for top spot, collapsed into this:

That last-second kick from Naitanui smashed the Giants' top-2 chances down to 1% and their top-4 chances to 27%.

It's now curiously difficult for Hawthorn to finish 2nd, as they'll be #1 if they win both games and probably 3rd or below if they don't. The real obstacle is West Coast in Perth this week: win that and they'll roughly double their amount of real estate at the top of the Tower.

Meanwhile Melbourne struck a wedge in 8th spot, expanding their finals chances from 0.4% to 6%, as the other main challenger, St. Kilda, were obliterated by the Swans.

Ladder Predictor! The squiggle is tipping the Eagles over the Hawks, which would see things end up something like this:

Flagpole! A small break on the rest of the comp for Sydney and Adelaide, with the Cats drifting.

Live squiggling!

I just wanted to say, you really are a talented person, Final Siren. Well done for all the squiggly goodness and insight you've brought to us for the last three years.

 

[Dogmatix]

Am I misunderstanding the Tower of Power, or is it giving Port a tiny chance of winning 3 out of its next 2 games and making the top 8?

You're misunderstanding it, the only point the matters is the midpoint of each position where the line direction changes

 

[RW]

Losing another GF by a point... geez...

If that happened, I'd be gutted, but proud of the team for what has been a brilliant season.

 

[Final Siren]

You're misunderstanding it, the only point the matters is the midpoint of each position where the line direction changes

This is correct. It's not very intuitive. It looks good earlier in the year because you get these nice organic shapes, but around now, as everything straightens out, there are little slivers here and there that suggest tiny chances where none exist.

 

[Final Siren]

I just wanted to say, you really are a talented person, Final Siren. Well done for all the squiggly goodness and insight you've brought to us for the last three years.

Thanks Rotayjay! It's my pleasure. Good fun.

 

[Final Siren]

Little question Final Siren , and I don't really need a full on explanation or detailed statistical breakdown but moreso just your opinion, but how would Geelong look on the squiggle and flagpole if we didn't drop those games to lesser teams?
I feel considering the Cats record vs top 8 sides that it should be higher, no doubt brought down by those crappy loses.
Would our position improve all that much?

Yes, it would. Squiggle cares a lot about results against poor sides.

Actual wins don't matter much, though. Beating Richmond by 4pts is about the same as losing to St Kilda by 3pts, from a squiggle perspective. What matters is that the Cats were evenly matched in those games, rather than winning comfortably.

 

[Final Siren]

North play GWS & Sydney. According to the ladder predictor they will win 1 of 2, but predictor is also saying that Swans and Giants both win both of theirs.

Am I missing something?

Just that the Ladder Predictor sacrifices the guarantee of producing a ladder that can actually exist in order to produce something that's pretty accurate on average.

It's a bit of a mind-bender, but the key concept is that instead of picking one particular outcome out of the multitude of possibilities, it's instead distilling all those into an average. That means leaving reality behind to a degree, because you run into fun paradoxes such as that Sydney and Hawthorn are both more likely to finish 1st than any other rung in particular. Or that a team is more likely to finish above another despite also being less likely to win the same number of games.

Basically you can't really represent a cloud of possibilities with a single ladder. But you have to consider the cloud if you want to make an accurate prediction! You can't just pick the single most likely outcome, because with 18 teams, there are 6,402,373,705,728,000 different ways to order them, so the most likely one will still be incredibly unlikely.

 

[Happy_Neale]

Just that the Ladder Predictor sacrifices the guarantee of producing a ladder that can actually exist in order to produce something that's pretty accurate on average.

It's a bit of a mind-bender, but the key concept is that instead of picking one particular outcome out of the multitude of possibilities, it's instead distilling all those into an average. That means leaving reality behind to a degree, because you run into fun paradoxes such as that Sydney and Hawthorn are both more likely to finish 1st than any other rung in particular. Or that a team is more likely to finish above another despite also being less likely to win the same number of games.

Basically you can't really represent a cloud of possibilities with a single ladder. But you have to consider the cloud if you want to make an accurate prediction! You can't just pick the single most likely outcome, because with 18 teams, there are 6,402,373,705,728,000 different ways to order them, so the most likely one will still be incredibly unlikely.

For real? No wonder betting companies run 'pick the ladder for a million bucks' competitions.

 

[Montana2Rice]

For real? No wonder betting companies run 'pick the ladder for a million bucks' competitions.

I'd assume final siren is correct but if you want to check him, get the calculator app on your phone or device and just go
18*17*16*15*14*13*12*11*10*9*8*7*6*5*4*3*2. (Or 18 and then hit the x! Button) You will need to turn your phone so you see the display in landscape.

It's simply saying, there is a possibility that any of the 18 teams can come first, then there is obviously 17 teams left so they could all come second, which leaves 16 teams left that could come 3rd, etc.

In reality, before the season, you could have made an educated guess that perhaps blues, lions, bombers, and a couple more could not come above top 12. So you could then narrow down the odds slightly. And the more educated guesses you make, the further you can narrow the odds. But doing that sees you lose out if a team like Leicester win the epl as happened last year, when all the experts were tipping them to be bottom 4, and a team like Chelsea who were unanimously tipped to be top 4, finish about 10th. Both in the same season totally screws you if you try to make an educated guess. The same way that the dockers finishing so low this year has probably screwed up any number of ladder predictions.

 

[philohk]

I don't mean to underrate other teams' chances, as I think it's pretty clear to all that there are a few teams this year that can win the flag. But I can't help thinking Flagpole is being skewed by the race for percentage between the teams just below us, which is forcing them to play at 100% against the lower teams rather than doing enough to win then putting the cue in the rack, whereas we're coasting a bit more. Or maybe it's rating us exactly where we are...

 

[jjsmitty8]

Just did my ladder predictor, this is what I've come up with. Don't shoot me, purely an opinion

Round 22 -
West Coast vs Hawthorn (Hawks very capable, but hoping the home ground advantage works fo the Eagles)
North vs Sydney (Sydney are red hot, North are not)
Richmond vs Saints (Better form, if it was at Etihad I wouldn't hesitate. 50-50 game depends on which sides actually turn up)
Giants vs Freo ($1.01 would be generous odds for GWS)
Magpies vs Gold Coast (A bit like Rich/St.K, depends who comes to play)
Port vs Adelaide (Too strong for their hometown rivals)
Carlton vs Melbourne (Better form and have something on the line, Melbourne of old would capitulate, they seem to have turned over a new leaf)
Lions vs Cats (They should win, even though they like to seem incompetent against lesser sides)
Bombers vs Dogs (Should be too good)

Round 23-
Adelaide vs West Coast (Just can't tip against Adelaide at home)
Geelong vs Melbourne (Melbourne rolled the Cats last year at Simmonds, would have to emulate that. Cats the superior team, SHOULD win)
Bombers vs Carlton (Hate Brisbane and Carlton, hope Essendon sours Carlton's season and sends Brisbane to last on %. Carlton should win though)
Sydney vs Richmond (Like I said above, Swans are hot, and Tigers are not. Sydney might want to get a little bit of payback for earlier in the year too)
Gold Coast vs Port (At Metricon, so i'll pick GC)
North v Giants (I think North will be safe regardless of the result, if GWS bring their A-game, they should be too good)
Saints v Lions (Saints are Etihad are a different side. Lions are deplorable everywhere. I have this win making them finish above Melbourne)
Hawks v Magpies (Hawks should tune up for finals)
Freo vs Dogs (Can't even make a case for home ground advantage. Absolute howler of a season for the Dockers.)

So that leaves my ladder looking like:

1. Sydney 17-5 148.4%
2. Adelaide 17-5 143.2%
3. Geelong 17-5 134.3%
4. Hawthorn 17-5 120%
5. GWS 16-6 138.4%
6. Bulldogs 16-6 116.3%
7. West Coast 15-7 125.9%
8. North Melbourne 12-10 105.3%
----------------------------------------
9. St.Kilda 12-10 95.3%
10. Melbourne 11-11 105.2%
11. Port Adelaide 9-13 103.3%
12. Collingwood 9-13 92.2%
13. Richmond 8-14 81.1%
14. Gold Coast 7-15 81.4%
15. Carlton 7-15 78.9%
16. Fremantle 3-19 74.8%
17. Brisbane 3-19 62.7%
18. Essendon 2-20 60.5%

Qualifying and Elimination Finals:
Sydney (1) vs Hawthorn (4) - Hawks have done it all before, Sydney at home is a huge ask. Sydney for me.
Adelaide (2) vs Geelong (3) - Yes, Geelong beat Adelaide in Adelaide earlier this year. Capable of a repeat, but like I said, can't tip against them at home.

Giants (5) vs North (8) - One meeting pre-bye, one post bye. Think GWS will win, AGAIN and it'll be curtains for North. The venue will be interesting.
Dogs (6) vs Eagles (7) - Again, venue will be interesting. Dogs forced to host a home final against an interstate side last year the G. No one jumps out at me, but the West Coast are generally pretty incompetent away from home.

Semi Finals:
Hawthorn vs Giants - Giants smashed them earlier in the year albeit at home. Hawks find another gear in September. Should be a ripper, but think Hawks will be too good.
Cats vs Dogs - Cats have beaten them twice throughout the year. I think they'll be successful this time as well.

Preliminary Finals:
Adelaide vs Hawthorn - I didn't think Hawks would beat Freo at Subi in the prelims last year. Don't think they can beat Adelaide at AO in the prelims this year.
Sydney vs Geelong - It seems to be a common trend for my tipping. I find it super hard to tip against quality interstate sides on their home deck.

Grand Final:
Sydney vs Adelaide - Adelaide have got the firepower to be strong enough on GF day. BUT, Sydney have been there before and they know what it takes. Their engine room takes over and Buddy lights up. Will be a cracking GF if it pans out this way, an super attacking and fierce game. Sydney gets it done in my opinion.

In summary, as much as I'd like it to, I don't think the Melbourne fairytale comes to fruition and North hang on bye the skin of their teeth despite not winning again this year. I genuinely think Melbourne's loss to Essendon in round 2 is going to cost them a place. They'd be in the 8 this weekend if they beat Carlton and North lose to Sydney.

If the Hawks win this Friday, it throws my whole prediction out of whack. If they lose, then it sets up a much harder run in the finals than if they win and finish top. They had a hard road last year, games against West Coast and Freo in Perth, if they lose this week and have to play Adelaide and Sydney away, I don't think they can win it.

I think there's only 5 teams who can genuinely go all the way, every team can be exposed, but Adelaide and Sydney are in the best form and look to possibly have the easiest road to the big dance.

 

[GreyCrow]

Animated!​

 

[Gralin]

I don't mean to underrate other teams' chances, as I think it's pretty clear to all that there are a few teams this year that can win the flag. But I can't help thinking Flagpole is being skewed by the race for percentage between the teams just below us, which is forcing them to play at 100% against the lower teams rather than doing enough to win then putting the cue in the rack, whereas we're coasting a bit more. Or maybe it's rating us exactly where we are...

I'd be careful about the cue in the rack analogy, freo supporters were adamant about it last year and we all know how that turned out

 

[cryptor]

Just did my ladder predictor, this is what I've come up with. Don't shoot me, purely an opinion

Round 22 -
West Coast vs Hawthorn (Hawks very capable, but hoping the home ground advantage works fo the Eagles)
North vs Sydney (Sydney are red hot, North are not)
Richmond vs Saints (Better form, if it was at Etihad I wouldn't hesitate. 50-50 game depends on which sides actually turn up)
Giants vs Freo ($1.01 would be generous odds for GWS)
Magpies vs Gold Coast (A bit like Rich/St.K, depends who comes to play)
Port vs Adelaide (Too strong for their hometown rivals)
Carlton vs Melbourne (Better form and have something on the line, Melbourne of old would capitulate, they seem to have turned over a new leaf)
Lions vs Cats (They should win, even though they like to seem incompetent against lesser sides)
Bombers vs Dogs (Should be too good)

Round 23-
Adelaide vs West Coast (Just can't tip against Adelaide at home)
Geelong vs Melbourne (Melbourne rolled the Cats last year at Simmonds, would have to emulate that. Cats the superior team, SHOULD win)
Bombers vs Carlton (Hate Brisbane and Carlton, hope Essendon sours Carlton's season and sends Brisbane to last on %. Carlton should win though)
Sydney vs Richmond (Like I said above, Swans are hot, and Tigers are not. Sydney might want to get a little bit of payback for earlier in the year too)
Gold Coast vs Port (At Metricon, so i'll pick GC)
North v Giants (I think North will be safe regardless of the result, if GWS bring their A-game, they should be too good)
Saints v Lions (Saints are Etihad are a different side. Lions are deplorable everywhere. I have this win making them finish above Melbourne)
Hawks v Magpies (Hawks should tune up for finals)
Freo vs Dogs (Can't even make a case for home ground advantage. Absolute howler of a season for the Dockers.)

So that leaves my ladder looking like:

1. Sydney 17-5 148.4%
2. Adelaide 17-5 143.2%
3. Geelong 17-5 134.3%
4. Hawthorn 17-5 120%
5. GWS 16-6 138.4%
6. Bulldogs 16-6 116.3%
7. West Coast 15-7 125.9%
8. North Melbourne 12-10 105.3%
----------------------------------------
9. St.Kilda 12-10 95.3%
10. Melbourne 11-11 105.2%
11. Port Adelaide 9-13 103.3%
12. Collingwood 9-13 92.2%
13. Richmond 8-14 81.1%
14. Gold Coast 7-15 81.4%
15. Carlton 7-15 78.9%
16. Fremantle 3-19 74.8%
17. Brisbane 3-19 62.7%
18. Essendon 2-20 60.5%

Qualifying and Elimination Finals:
Sydney (1) vs Hawthorn (4) - Hawks have done it all before, Sydney at home is a huge ask. Sydney for me.
Adelaide (2) vs Geelong (3) - Yes, Geelong beat Adelaide in Adelaide earlier this year. Capable of a repeat, but like I said, can't tip against them at home.

Giants (5) vs North (8) - One meeting pre-bye, one post bye. Think GWS will win, AGAIN and it'll be curtains for North. The venue will be interesting.
Dogs (6) vs Eagles (7) - Again, venue will be interesting. Dogs forced to host a home final against an interstate side last year the G. No one jumps out at me, but the West Coast are generally pretty incompetent away from home.

Semi Finals:
Hawthorn vs Giants - Giants smashed them earlier in the year albeit at home. Hawks find another gear in September. Should be a ripper, but think Hawks will be too good.
Cats vs Dogs - Cats have beaten them twice throughout the year. I think they'll be successful this time as well.

Preliminary Finals:
Adelaide vs Hawthorn - I didn't think Hawks would beat Freo at Subi in the prelims last year. Don't think they can beat Adelaide at AO in the prelims this year.
Sydney vs Geelong - It seems to be a common trend for my tipping. I find it super hard to tip against quality interstate sides on their home deck.

Grand Final:
Sydney vs Adelaide - Adelaide have got the firepower to be strong enough on GF day. BUT, Sydney have been there before and they know what it takes. Their engine room takes over and Buddy lights up. Will be a cracking GF if it pans out this way, an super attacking and fierce game. Sydney gets it done in my opinion.

In summary, as much as I'd like it to, I don't think the Melbourne fairytale comes to fruition and North hang on bye the skin of their teeth despite not winning again this year. I genuinely think Melbourne's loss to Essendon in round 2 is going to cost them a place. They'd be in the 8 this weekend if they beat Carlton and North lose to Sydney.

If the Hawks win this Friday, it throws my whole prediction out of whack. If they lose, then it sets up a much harder run in the finals than if they win and finish top. They had a hard road last year, games against West Coast and Freo in Perth, if they lose this week and have to play Adelaide and Sydney away, I don't think they can win it.

I think there's only 5 teams who can genuinely go all the way, every team can be exposed, but Adelaide and Sydney are in the best form and look to possibly have the easiest road to the big dance.

For your prediction to come off it requires Hawthorn to have gone:

Loss (Melbourne)
Win (North Melbourne)
Loss (West Coast)
Win (Collingwood)
Loss (Sydney)
Win (GWS)
Loss (Adelaide)

That's just 3 wins from 7 games. Or to put it another way, 3 losses from our next 5 games. We've only had 3 losses from our past 15 games.

I understand Hawthorn are on a decline and I'm not saying your prediction can't happen... but geez, that'd be a sudden drop off. Particularly with so many proven finals performers fit and playing good footy. And with blokes like Hodge and Burgoyne still a possibility of retiring at the end of the season I can't see them not responding strongly to a loss like we did on the weekend.

Does highlight at least just how important it is for Hawthorn to win their next 2 games and lock up top spot for home finals.

 

[wallyt99]

As a hawks fan I reluctantly agree with Flagpole... But does Flagpole take into account likely ( but hugely unfair according to Freo, Sydney and West Coast fans) home ground advantage?

If we win our last couple - which is as I understand it predicted - then surely our flag chances improve dramatically with the chance of 3 games at the G.

 

[cryptor]

As a hawks fan I reluctantly agree with Flagpole... But does Flagpole take into account likely ( but hugely unfair according to Freo, Sydney and West Coast fans) home ground advantage?

If we win our last couple - which is as I understand it predicted - then surely our flag chances improve dramatically with the chance of 3 games at the G.

We're predicted to lose this week by a goal. If we win the predictions should change to be a lot more favourable of us.

 

[The King!]

Hawks will win the last two games comfortably

 

[MarcusP2]

Am I misunderstanding the Tower of Power, or is it giving Port a tiny chance of winning 3 out of its next 2 games and making the top 8?

Looks like an artifact of using shapes instead of bars. Chance is 0 at 8th and some sliver at 9, so the slope shows up. Poor data presentation Final Siren, they're discrete data points not a continuum!

 

[Ticky009]

- then surely our flag chances improve dramatically with the chance of 3 games at the G.

Ooh, don't think you really want to raise that bug bear

 

[chazwazza]

Totally erratic and pointless question Final Siren but I look at a few different measures and probabilities concerning outcomes this year. One of them is the Premiership Standard
http://premiershipstandard.com/?rnd=current
A while ago you posted a pic of "The Squiggle Sweet Spot" which was the centre of all premierships. I just wonder if this might also fall into the centre of the Premiership Quadrant. I've no idea how you would determine this and just thinking out aloud so apologies if it's a stupid and irrational question.

 

[Hobbes]

For real? No wonder betting companies run 'pick the ladder for a million bucks' competitions.

Even if you know which teams will make up the top eight, and which teams will make up the bottom ten, there are still 146313216000 possibilities.