Certified Legendary Thread Race for the flag, in squiggly lines

[Macca81]

squiggle?

 

[Suntory Boss Coffee]

Round 12

Adelaide 85 - 102 Hawthorn

Adelaide Oval (SA)0.31 wins - 0.69 wins

Richmond 79 - 75 West Coast

M.C.G. (VIC)0.55 wins - 0.45 wins

Carlton 77 - 104 Port Adelaide

M.C.G. (VIC)0.21 wins - 0.79 wins

Greater Western Sydney 91 - 90 North Melbourne

Sydney Showground (NSW)0.53 wins - 0.47 wins

Western Bulldogs 104 - 71 Brisbane Lions

Docklands (VIC)0.85 wins - 0.15 wins

Geelong 108 - 55 Melbourne

Kardinia Park (Gee)0.97 wins - 0.03 wins

 

[KillaKanga]

Round 12

Greater Western Sydney 91 - 90 North Melbourne

Sydney Showground (NSW)0.53 wins - 0.47 wins

Betting Angencies say North to win. Squiggle says Giants.

Guess I'll wait til 3 quarter time. If North are up, they'll win. Right? #4quarterspecialists

 

[SJ]

Betting Angencies say North to win. Squiggle says Giants.

Guess I'll wait til 3 quarter time. If North are up, they'll win. Right? #4quarterspecialists

Squiggle doesn't know about the injury decimation that GWS has suffered in the past few days.

 

[Richo83]

Yeah injuries are Roby's thing. Here's the thing, I agree that injuries pretty much dictate that North should be favorites, and they should win. But Roby gets criticized for being too subjective in his injury assessments.

 

[krisholio14]

I'm not even trolling when I say this, it sucks for Fremantle they don't have a premiership because their movements can't be compared to precedent. They could do a figure eight and no one would know what that means.

Not against their own premiership precedent, but that's the whole idea of other premiers being on the chart. It gives you an idea of where they are at relative to other recent premiership sides.

This much we know. Ross Lyon sides always find themselves somewhere around the two Sydney premierships. Which, surprise surprise, means they are on the ultra defensive side of the premiership curve. Not too many teams win from out there.

 

[Heeney2Franklin]

Not against their own premiership precedent, but that's the whole idea of other premiers being on the chart. It gives you an idea of where they are at relative to other recent premiership sides.

This much we know. Ross Lyon sides always find themselves somewhere around the two Sydney premierships. Which, surprise surprise, means they are on the ultra defensive side of the premiership curve. Not too many teams win from out there.

It does paint a picture that flags of the last 20 years are weighted at 70% and above for attack.

Not sure what it means in the grand scheme of things.

 

[krisholio14]

It does paint a picture that flags of the last 20 years are weighted at 70% and above for attack.

Not sure what it means in the grand scheme of things.

It means Fremantle (and Ross Lyon) will be a bit of an outlier if they do win it this year.

 

[cryptor]

It does paint a picture that flags of the last 20 years are weighted at 70% and above for attack.

Not sure what it means in the grand scheme of things.

Given the Hawks are currently the only side above 70 for attack hopefully it means it's a 1 horse race

 

[Heeney2Franklin]

Given the Hawks are currently the only side above 70 for attack hopefully it means it's a 1 horse race

Well look at 2013 you had the weakest defence of the top 4 sides but it didn't matter because no matter what you conceded you would kick over 100pts almost every time.

It is fascinating tho why not many flags are won on the opposite side of the spectrum and generally favor all attacking teams.

It's also why your still flag favorites with the punters.

 

[krisholio14]

Well look at 2013 you had the weakest defence of the top 4 sides but it didn't matter because no matter what you conceded you would kick over 100pts almost every time.

It is fascinating tho why not many flags are won on the opposite side of the spectrum and generally favor all attacking teams.

It's also why your still flag favorites with the punters.

It flies in the face of the generally accepted philosophy that defense wins flags.

 

[cryptor]

Well look at 2013 you had the weakest defence of the top 4 sides but it didn't matter because no matter what you conceded you would kick over 100pts almost every time.

It is fascinating tho why not many flags are won on the opposite side of the spectrum and generally favor all attacking teams.

It's also why your still flag favorites with the punters.

You can't win games just by preventing the opposition from scoring. You ultimately have to score to win and the team that scores best logically will win if everything else is equal. The Squiggle has that combined offence and defense rating but in practice I think the offence has a bit more of an advantage.

I think that because scoring is something particularly potent that can change a game in a matter of minutes (eg. Hawthorn's ability to go for 15 minutes in a quarter without scoring and then in the space of 5 minutes pile on a match winning lead) where as defense is usually a grind and against a good scoring side is something that can't be lapsed on if you want to win.

 

[Final Siren]

squiggle?

Sorry, I forget there's football on when Richmond aren't playing. Well not really. I will squiggle tomorrow.

 

[The 747]

It flies in the face of the generally accepted philosophy that defense wins flags.

It's more a condition, you generally need a good defence to be anywhere near a flag. Bad defence against a top 2 side isn't going to end well generally.

 

[Rotayjay]

You can't win games just by preventing the opposition from scoring. You ultimately have to score to win and the team that scores best logically will win if everything else is equal. The Squiggle has that combined offence and defense rating but in practice I think the offence has a bit more of an advantage.

I think that because scoring is something particularly potent that can change a game in a matter of minutes (eg. Hawthorn's ability to go for 15 minutes in a quarter without scoring and then in the space of 5 minutes pile on a match winning lead) where as defense is usually a grind and against a good scoring side is something that can't be lapsed on if you want to win.

Long live attacking football!

 

[Final Siren]

You can't win games just by preventing the opposition from scoring. You ultimately have to score to win and the team that scores best logically will win if everything else is equal. The Squiggle has that combined offence and defense rating but in practice I think the offence has a bit more of an advantage.

This seems to be the case for flags. Here is the last 20 years of flag winners vs all other teams (season finishing position):

So first let's draw a line at the point of the lowest-ranked premiership (Geelong 2009), signifying the minimum level a team has needed to reach to win a flag since 1995. The line keeps constant the total of attack + defence.

This gives us two zones:

Zone | Flags | Misses | Flag % \ Above the line | 20 | 27 | 43% \ Below the line | 0 | 280 | 0%

Now let's add a line separating teams into more attacking vs more defensive.

Teams in the zone marked "A" have higher attack scores than defence scores, and "D" are the reverse.

Zone | Flags | Misses | Flag % \ A (more attacking) | 15 | 13 | 54% \ D (more defensive) | 5 | 14 | 26%

So clearly the attacking area of the chart looks more fruitful: if you finished at least as high as Geelong 2009, you won a flag 54% of the time if you were an attacking team, but only 26% of the time if you were a defensive team.

And this difference becomes more pronounced if you section off the teams that only just qualified. For example, if I add an arbitrary line that excludes teams within 8 points of the intersection:

Zone | Flags | Misses | Flag % \ A1 (very attacking) | 10 | 3 | 77% \ A2 (mildly attacking) | 5 | 10 | 33% \ D1 (very defensive) | 2 | 4 | 33% \ D2 (mildly defensive) | 3 | 10 | 23%

The sample size here isn't huge, but the numbers are pretty compelling. If you're good enough to challenge for the flag, you've been a lot better off being attacking.

 

[Tazmania]

POTY contender right there folks.

 

[Macca81]

POTY contender right there folks.

Not really. I was hoping to see west coast move up and to the right after thumping essendon by 50 points

 

[MF]

Mr Squiggle... Do you have a tipping algorithm that gives weighting to where a team is in the attacking/defensive spectrum?

 

[cryptor]

Zone | Flags | Misses | Flag % \ A1 (very attacking) | 10 | 3 | 77% \ A2 (mildly attacking) | 5 | 10 | 33% \ D1 (very defensive) | 2 | 4 | 33% \ D2 (mildly defensive) | 3 | 10 | 23%

The sample size here isn't huge, but the numbers are pretty compelling. If you're good enough to challenge for the flag, you've been a lot better off being attacking.

Given ISTATE-91:12 struggles with the GF tip, have you tried using this particular set of data to do that job?

 

[Final Siren]

Round 11, 2015​

Animated!

Good week for: Geelong again! The Cats have come a long way since Round 4, when their solitary win was courtesy of a 9-point margin (at home!) over Gold Coast. Since then, they've dispatched Richmond, Collingwood, Carlton, Essendon, and Port Adelaide, with losses only against West Coast and Sydney, both away. The squiggle says that's pretty good.

Not too shabby for the Pies, either, who have also travelled a long way since the start of the year. The squiggle still rates them lower than their current ladder position would suggest -- and wasn't quite as excited about their win over GWS as the media -- but they're on track for finals at the moment.

Bad week for: Fremantle, who are getting a bit of an alarming slide on. In the three weeks since Round 8, when they pulverised North Melbourne, the Dockers have turned in unconvincing games against Adelaide, Richmond, and a severely weakened Gold Coast.

As a result, the ladder predictor now has Sydney on top! Although only just:

North Melbourne, who used to torment the squiggle with wildly unpredictable performances, turned in yet another game just like the squiggle expected. It's almost eerie now how. So once again they don't squiggle very far.

Neither do Sydney, nor St Kilda and Melbourne, who also played out within a few points of prediction. And West Coast belted Essendon by 8 goals, but were expected to by 7, so they're another two teams who pretty much stay where they are.

Interactive squiggle, FAQ, and the like

 

[Final Siren]

Given ISTATE-91:12 struggles with the GF tip, have you tried using this particular set of data to do that job?

I'll take another look at Grand Final tipping later this year. It's always a bit dodgy, building a prediction out of data when you only get one sample a year (two if the ball bounces the wrong way), but yes, I am tempted to include an attacking bias now. Not for regular games, just GFs.

 

[Membling]

Wouldn't mind swans to trek north a bit...

 

[wcadams88]

With the remaining games for the Cats, what would need to be seen for them to treck towards cup Territory with a predicted 1 game out of the 45

Sent from my SM-N915G using Tapatalk

 

[krisholio14]

It's more a condition, you generally need a good defence to be anywhere near a flag. Bad defence against a top 2 side isn't going to end well generally.

What is interesting is that if you go back and have a look at the squiggle over the years, very rarely til the 70's did strong attacking teams win it.

So once upon a time defense did win flags, because most of the premierships pre 1970 were won out in Sydney 05/12 territory or beyond.

Look at the difference between Essendon's 49/50 premiership sides and Geelong's 51/52 premiership sides as an example. Especially Essendon 1950. For a team from 65 years ago to spend the whole season in the premiership sweet spot of today...geez that must have been one hell of a team (and that John Coleman one hell of a player, among others).

I think it just goes to show how the game has evolved due to improved fitness, skill levels and playing conditions.