Certified Legendary Thread Race for the flag, in squiggly lines

[Final Siren]

I was just on afl tables and having a look at Adelaide records and saw their 2009 run of good defensive games.
24 to North in Rd 12, then 13 to Freo in Rd 15.
So I went to look at the squiggle for 2009. And it surprised me. It appears to the naked eye that Adelaide was the best side that year.
Is that right?

They finished the year in a better position, yes! Which is very unusual for a team that didn't even make a prelim, and probably doesn't match up with general wisdom of how good they really were.

So what happened was the Crows were well behind St Kilda and Geelong for most of the year, but had a huge finish, and got ahead in Round 22 when they thrashed Carlton (7th) by 76 points. Then they beat Essendon by 16 goals in their first final. So at this point the squiggle was very impressed, whereas it had never been that taken with Geelong that year, who tended to fall over the line in games against weak opposition. (The Cats final percentage was only 127%.)

Then the Crows lost by 5 points to Collingwood at the MCG in a semi-final and that was that.

 

[Dirty Bird]

They finished the year in a better position, yes!

How many times in history has this happened?
I mean, the highest rated team according the squiggle failing to make the final 4?

 

[Dirty Bird]

Our % went up by too much.
Cost us a semi according to the squiggle.

 

[cryptor]

Our % went up by too much.
Cost us a semi according to the squiggle.

You'll just have to enjoy the semi you got from today's result instead.

 

[Catsace]

Bump, simply because this great thread needs staying on the front page.

Guessing too with Geelong's relatively fortuitous win over Freo, Geelong moved to the left quite a bit, while the Dockers stayed about the same. Guessing the Squiggle wanted a bit more of a convincing effort. Will be fascinating this week, the Hawks vs Freo, almost the battle of ideological opposites.

Also interesting to see the Tiger start to close into the action, the Squiggle must think it's almost Tiger Time!

 

[Final Siren]

Round 20, 2014​

This week's big mover is Adelaide! Brisbane have been keeping sides to low scores lately: going backwards from Round 19 it's been Melbourne 51, Gold Coast 56, West Coast 76, Richmond 79, North Melbourne 66. So the squiggle is impressed that the Crows came out and kicked 168 against them (at the Gabba).

As Dirty Bird noted, this performance was in fact so impressive that the squiggle ladder predictor bumped them from 8th to 7th, which means it no longer expects a Port vs Adelaide Elimination Final, which means the Crows will have to travel, which means it thinks they will lose. So, bad move, Adelaide.

Collingwood is nice and easy to see this week, for those who were concerned about that.

The top 4 didn't move a whole lot, but Sydney had the best of it.

I've been keeping an eye on the interactive squiggle ladder predictor and I'm pretty happy with how it's gone. It seems nice and stable; for example, it has been resolutely slotting Richmond into 12th place since Round 11, when the Tigers were 3-7, keeping them there as they slid to 3-10 and recovered to 9-10. So while common wisdom had the Tigers plunging and then soaring, the predictor wasn't moved. It may still be wrong, of course, if Richmond can pull off some history-defying heroics, but in terms of making cold-hearted, conservative predictions based on current form, I think the predictor does it right.

It is a little quirky, in that it predicts ladders that aren't quite possible: The only way for 7th to 12th to finish in the order currently predicted is if some of the teams don't win the number of games predicted. But such is its nature.

Looking back, here is a comparison with the actual ladder.

After Round 5:

# | AFL Ladder | Squiggle Predictor
\1|Geelong|Geelong
\2|Port Adelaide|Hawthorn ↑1
\3|Hawthorn|Port Adelaide ↓1
\4|West Coast|Fremantle ↑1
\5|Fremantle|Collingwood ↑1
\6|Collingwood|North Melbourne ↑1
\7|North Melbourne|Richmond ↑4
\8|Gold Coast|Adelaide ↑4
\9|St Kilda|Essendon ↑1
\10|Essendon|Sydney ↑3
\11|Richmond|West Coast ↓7
\12|Adelaide|Gold Coast ↓4
\13|Sydney|Western Bulldogs ↑2
\14|GWS|St Kilda ↓5
\15|Western Bulldogs|Carlton ↑1
\16|Carlton|Brisbane ↑2
\17|Melbourne|GWS ↓3
\18|Brisbane|Melbourne ↓1

That looks pretty reasonable to me, since in almost all cases, the squiggle shifted teams towards where they now look like finishing.

And after Round 11:

# | AFL Ladder | Squiggle Predictor
\1|Port Adelaide|Sydney ↑2
\2|Hawthorn|Port Adelaide ↓1
\3|Sydney|Hawthorn ↓1
\4|Collingwood|North Melbourne ↑4
\5|Gold Coast|Geelong ↑1
\6|Geelong|Fremantle ↑1
\7|Fremantle|Collingwood ↓4
\8|North Melbourne|Essendon ↑2
\9|Adelaide|Gold Coast ↓4
\10|Essendon|Adelaide ↓1
\11|West Coast|West Coast
\12|Carlton|Richmond ↑1
\13|Richmond|Carlton ↓1
\14|Western Bulldogs|Western Bulldogs
\15|Melbourne|Melbourne
\16|St Kilda|Brisbane ↑2
\17|GWS|St Kilda ↓1
\18|Brisbane|GWS ↓1

 

[Final Siren]

How many times in history has this happened?
I mean, the highest rated team according the squiggle failing to make the final 4?

I have been fiddling with algorithms a bit so I'm not sure this lines up 100% with the online squiggles, but: in the last 20 years, only that one time.

Going backwards from 2013, the team that finished "highest" in the squiggle (by simply adding OFFENCE + DEFENCE) finished: 1st, 2nd, 1st, 1st, 5th (Adelaide 2009), 2nd, 1st, 1st, 3rd (Adelaide 2005), 1st, 1st, 1st, 1st, 1st, 4th (Brisbane 1999), 1st, 2nd, 1st, 1st, 1st.

 

[Dylan8]

I'm finding it very hard to take reading the squiggle in the last month. However, I do acknowledge it is the ultimate objective power.

 

[Zuta]

If Geelong were to finish 2nd and Hawks 3rd who would the home ground advantage be awarded to? Does the Squiggle just assume the game is at Simmonds or does it know its at the MCG and award the Hawks the advantage?

 

[Final Siren]

If Geelong were to finish 2nd and Hawks 3rd who would the home ground advantage be awarded to? Does the Squiggle just assume the game is at Simmonds or does it know its at the MCG and award the Hawks the advantage?

It will blindly award advantage to whoever finishes 2nd.

At the start of this year, I didn't expect this kind of interest in the squiggle, and I configured it with the simplest algorithm. Unfortunately that means it does dumb things with home advantage sometimes, especially with Geelong. I will probably change that for 2015. It doesn't make a big difference overall, but it's confusing when you're scrutinizing individual games, like finals involving Geelong.

 

[Blue and Silver]

It's nice to take a look at this thing and see some progress for Carlton.

Got to start somewhere

 

[TheWoodenSlug]

It will blindly award advantage to whoever finishes 2nd.

At the start of this year, I didn't expect this kind of interest in the squiggle, and I configured it with the simplest algorithm. Unfortunately that means it does dumb things with home advantage sometimes, especially with Geelong. I will probably change that for 2015. It doesn't make a big difference overall, but it's confusing when you're scrutinizing individual games, like finals involving Geelong.

Maybe the squiggle was predicting us to play Freo in the first final?

Oh, wait..

 

[Golumless]

I have been fiddling with algorithms a bit so I'm not sure this lines up 100% with the online squiggles, but: in the last 20 years, only that one time.

Going backwards from 2013, the team that finished "highest" in the squiggle (by simply adding OFFENCE + DEFENCE) finished: 1st, 2nd, 1st, 1st, 5th (Adelaide 2009), 2nd, 1st, 1st, 3rd (Adelaide 2005), 1st, 1st, 1st, 1st, 1st, 4th (Brisbane 1999), 1st, 2nd, 1st, 1st, 1st.

Wow... we have really under performed in finals...

 

[Billy ray]

At the start of this year, I didn't expect this kind of interest in the squiggle.

I think everyone underestimated the lure of the squiggle... Soooo.... you have any plans for Oct to March?
You might have to have yourself a huge off season to add all the bells and whistles you have probably thought about over the years, and people have raised this season.
Spending time with family and friends in offseason? Ain't nobody got time for dat!

 

[Headless]

Carlton 86 - 94 Geelong

(8 points).

Final score: 76 - 82 (6 points).

 

[B4Bear]

Carlton 86 - 94 Geelong

(8 points).

Final score: 76 - 82 (6 points).

All hail the squiggle.

 

[Billy ray]

Squiggle did alright too on margins and basic magnitude of swans v saints and gc v port. And with 10 mins to go, ess v west Coast looking good too.

 

[CharlieG]

The Squiggle is now projecting West Coast to make the eight (despite losing today!) on just 11 wins. Strange season.

 

[The Swans Blog]

WHAT THE HELL HAPPENED???

Sorry for the caps, but Brisbane destroying Collingwood - what in the ****? Someone has got to explain this for me, I'm not sure how it could have possibly happened.

Looks like the Hawks will have a cake walk leading into the finals.

The Squiggle is now projecting West Coast to make the eight (despite losing today!) on just 11 wins. Strange season.

No one picked Collingwood to lose so much!

 

[FootyGeek]

Blues ahead of Pies on the squiggle now.

All that matters.

 

[The Speaker]

WHAT THE HELL HAPPENED???

Sorry for the caps, but Brisbane destroying Collingwood - what in the ****? Someone has got to explain this for me, I'm not sure how it could have possibly happened.

Lost Pendlebury before the match and Beams, Reid and Cloke during the match.

 

[Sydney Bloods]

just shows how tight the comp is nowadays.

10 years ago you take 4 players out of a strong team even players like beams, pendles, cloke and the reid against a side...... which lets be honest is pretty woeful and instead of a drumming we would have expected it to be a tight game, with the team with the depleted bench to fall away in the 4th and win by 30 odd points at best.

nowadays you lose your 4 best players to even a shit team it goes the other way and you get smashed.
which further highlights what an amazing squad hawthorn has, I can't find a team that covered injuries as well as they have this year.

 

[Final Siren]

The Squiggle is now projecting West Coast to make the eight (despite losing today!) on just 11 wins. Strange season.

Yes, a Bradbury for West Coast! Here is why: The squiggle sees it as more likely than not that each of the 4 teams ahead of the Eagles (Adelaide, Richmond, Collingwood, Gold Coast) will go 1-1 in their remaining games. Only West Coast is likely to win both games.

(I'm not sure if this is actually possible; Gold Coast may need to go 0-2. But that doesn't really matter.)

The reason is that to have a better than even chance of winning your next two games, your win likelihood for each game must average a little over 70%. Because 0.7 x 0.7 = 0.49.

When the win likelihood averages closer to 50%, e.g. when flipping coins, it's a lot more likely that you'll go 1-1 than 2-0. That is, you're 25% likely to go 2-0, 50% likely to go 1-1, and 25% likely to go 0-2.

The squiggle is cutting a fine line here between situations that are only slightly more likely than others, but that's the way it's calling it now!

 

[Final Siren]

Squiggle Public Service Announcement: I am travelling overseas at the moment so there will be no squiggle posted this week until maybe Thurs or Friday. The interactive squiggle will keep working though.

 

[Bobby_]

Lost Pendlebury before the match and Beams, Reid and Cloke during the match.

They were already getting smashed in the first quarter before Reid went off the ground and although it played a part, we have seen a similar scenario with Gold Coast on two occasions this year (last night against Port), and they still remained competitive.
Some of Collingwood's youth got exploited. Their rucks should have won the battle given it was 2 v 1 all night long but they didn't, got annihilated by Stefan Martin, while some of their leaders simply shut up shop for the night.