Certified Legendary Thread The Squiggle is back in 2023 (and other analytics)

[juss]

Power rankings after round 3

1. West Coast Eagles

2. Geelong Cats

3. GWS Giants

4. Collingwood Magpies

5. Brisbane Lions

6. Port Adelaide

7. Hawthorn

8. Richmond

9. Gold Coast

10. Bulldogs

11. Fremantle

12. Adelaide

13. St Kilda

14. Sydney Swans

15. Essendon

16. Melbourne

17. Carlton

18. North Melbourne

 

[blueiris]

Power rankings after round 3

1. West Coast Eagles

2. Geelong Cats

3. GWS Giants

4. Collingwood Magpies

5. Brisbane Lions

6. Port Adelaide

7. Hawthorn

8. Richmond

9. Gold Coast

10. Bulldogs

11. Fremantle

12. Adelaide

13. St Kilda

14. Sydney Swans

15. Essendon

16. Melbourne

17. Carlton

18. North Melbourne

The Lions should be higher we are the only team that has beaten 2 top 8 teams this season

 

[Viceregal]

The Lions should be higher we are the only team that has beaten 2 top 8 teams this season

Is there a post somewhere I can read about how the power rankings are worked out / what they represent? juss ?

 

[juss]

The Lions should be higher we are the only team that has beaten 2 top 8 teams this season

They will be, always rising.

Is there a post somewhere I can read about how the power rankings are worked out / what they represent? juss ?

Didnt do analysis this week these were thrown together this arvo. Next week will post the updated rankings and a brief sentence as to why.

For the record my rankings aren't based on a data model, they are just my personal rankings of the teams from best to worst, so early season is heavily weighted to my preseason rankings of a side before I think its definitely proven otherwise. They're far from perfect and I know that, but it's a bit of fun.

 

[nobbyiscool]

For the record my rankings aren't based on a data model, they are just my personal rankings of the teams from best to worst, so early season is heavily weighted to my preseason rankings of a side before I think its definitely proven otherwise. They're far from perfect and I know that, but it's a bit of fun.

Right. So we have a thread dedicated to squiggle and analytics... but you're such a self indulgent narcissist that you think your opinion is worth polluting the thread with?

 

[juss]

Right. So we have a thread dedicated to squiggle and analytics... but you're such a self indulgent narcissist that you think your opinion is worth polluting the thread with?

Wow, that's over the top lol.
It's a footy forum and I'm posting my own rankings to see how they compare with the others in this thread.

I'm very happy if you put my posts on ignore so they won't bother you

 

[Billy ray]

Right. So we have a thread dedicated to squiggle and analytics... but you're such a self indulgent narcissist that you think your opinion is worth polluting the thread with?

 

[Occidental]

Wow, that's over the top lol.
It's a footy forum and I'm posting my own rankings to see how they compare with the others in this thread.

I'm very happy if you put my posts on ignore so they won't bother you

Your feelings aren’t facts.
This is a thread about analytics.
Why don’t we just all post how we feel each round and then we can subjectively rank all our rankings?
This weekly opinion piece of yours is in the wrong thread.

 

[juss]

Your feelings aren’t facts.
This is a thread about analytics.
Why don’t we just all post how we feel each round and then we can subjectively rank all our rankings?
This weekly opinion piece of yours is in the wrong thread.

That's fine fair call, thank you for explaining it without the keyboard warrior psychoanalysis attached to it like the other poster.

 

[Final Siren]

From memory I think only state matters - so Hobart and Launceston are treated the same.

Both state and ground matters. Since last year, Squiggle uses a "Familiarity" method of establishing home ground advantage, which means the more often you've played at this ground and in this state compared to your opposition, the more HGA you have.

In reality, HGA is probably the result of several things, including crowd noise, umpiring, travel, fatigue, and psychology, which are all challenging to measure. But ground familiarity seems to be a pretty good working proxy for those.

 

[Final Siren]

I'm curious about this, as my team (Hawthorn) spent most of their dominant period "coasting" to wins (and even losses) during the H&A season.

Indeed it was a relatively common point of difference raised between Hawthorn and Geelong fans - Cats going at 100% always (more wins, more %), Hawks always keeping something in reserve (when it matters)....

Or are you suggesting Hawthorn fans were delusional and they really weren't better in finals than H&A during that period (say 2012-2014, with 15/16 being "one" last shot)

I don't recall much coasting from Hawthorn 2012-2015... the Hawks beat up on weak teams a lot. I do remember having the usual argument with people about Squiggle seemingly overrating these thumpings -- e.g. Round 17, 2015, when Hawthorn beat Carlton 173 to 35, Squiggle took that as an important sign of a powerful team, while many humans did not, because it was only Carlton.

The team I do remember as having a reputation for "coasting" to wins, and which seemed to bear it out, was Geelong 2009. The Cats won a ton of games (18) but often not by much, and there was a lot of talk about them having an extra gear to go to when it really mattered. I'm still not convinced that gear was there -- I think they were pretty fortunate to win some of those games -- but hey, they won the flag.

 

[Chaisa]

When was the last time Sydney were on bottom of the Squiggle rankings?

 

[Hobbes]

Sydney +2 v Melbourne
Collingwood +20 v Western Bulldogs
Geelong +14 v GWS
Essendon v Brisbane +6
Port Adelaide +9 v Richmond
North Melbourne v Adelaide +6
West Coast +24 v Fremantle
Gold Coast +19 v Carlton
St Kilda v Hawthorn +5

3/9. Possibly a bleak week for tipsters, or at least this tipster. Running total 18/36.



1. West Coast 22.8
2. Geelong 20.0
3. GWS 17.6 (+1)
4. Collingwood 17.3 (-1)
5. Essendon 4.9 (+7)
6. Fremantle 3.3 (+2)
7. Brisbane 1.3 (-2)
8. St Kilda 0.1 (+5)
9. Western Bulldogs -0.8 (+1)
10. Richmond -1.3 (+1)
11. Hawthorn -1.4 (-4)
12. Melbourne -3.15 (+2)
13. Port Adelaide -3.23 (-4)
14. Adelaide -3.3 (-8)
15. Gold Coast -7.0
16. North Melbourne-10.8 (+1)
17. Sydney -11.6 (-1)
18. Carlton -18.2

It's so close in the middle that Adelaide dropped four points, and eight places. I know it's early days, but four standout teams so far.

Brisbane v Collingwood +10
North Melbourne v Essendon +16
West Coast +39 v Port Adelaide
GWS +26 v Fremantle
Melbourne v St Kilda +3
Richmond +16 v Sydney
Western Bulldogs +17 v Carlton
Adelaide +14 v Gold Coast
Hawthorn v Geelong +21

 

[Forward Press]

I know it's early days, but four standout teams so far.

And we beat two of those four very convincingly.

Geelong in Geelong is probably the best barometer as to where we really sit this season.

 

[Hobbes]

A different team list. This list just rates each team's results for this year. It uses my algorithm to judge the quality of opponents, but doesn't give additional weight to the most recent results, and (apart from rating strength of oppos) doesn't take into account results before round 1.

Teams are listed by their average result (compared to a hypothetical average team playing the same matches).

1. Geelong 30.2
2. GWS 30.1
3. West Coast 22.6
4. Collingwood 14.3
5. Fremantle 13.5
6. Brisbane 9.4
7. St Kilda 5.2
8. Hawthorn 2.9
9. Essendon 1.8
10. Western Bulldogs 0
11. Port Adelaide -0.5
12. Gold Coast -1.5
13. Richmond -6.5
14. Adelaide -10.7
15. Sydney -18.9
16. Melbourne -19.5
17. Carlton -22.8
18. North Melbourne -23.3

 

[Suma Magic]

Final Siren

I tipped 6 upsets with the tipping system you helped shape. I got 7/9 and won the round in my 95-person work comp to get the weekly prize. Now just one point off first too.

Thanks for your help. I’d say that the system is a minor success (made slightly more money than the entry fee) even after 4 rounds.

 

[kaiserchief13]

Final Siren

I tipped 6 upsets with the tipping system you helped shape. I got 7/9 and won the round in my 95-person work comp to get the weekly prize. Now just one point off first too.

Thanks for your help. I’d say that the system is a minor success (made slightly more money than the entry fee) even after 4 rounds.

How could you tip 6 upsets when Collingwood, Essendon, WC and GC won?

 

[Suma Magic]

How could you tip 6 upsets when Collingwood, Essendon, WC and GC won?

Edit: actually only tipped 5 upsets and 4 faves pies eagles crows dons... Got 3/4 faves and 4/5 upsets I tipped.

 

[NoobPie]

Sydney +2 v Melbourne
Collingwood +20 v Western Bulldogs
Geelong +14 v GWS
Essendon v Brisbane +6
Port Adelaide +9 v Richmond
North Melbourne v Adelaide +6
West Coast +24 v Fremantle
Gold Coast +19 v Carlton
St Kilda v Hawthorn +5

3/9. Possibly a bleak week for tipsters, or at least this tipster. Running total 18/36.




6/9. Running total 15/27. And my first exact margin prediction of the year - thank you Brisbane.

1. West Coast 22.8
2. Geelong 20.0
3. GWS 17.6 (+1)
4. Collingwood 17.3 (-1)
5. Essendon 4.9 (+7)
6. Fremantle 3.3 (+2)
7. Brisbane 1.3 (-2)
8. St Kilda 0.1 (+5)
9. Western Bulldogs -0.8 (+1)
10. Richmond -1.3 (+1)
11. Hawthorn -1.4 (-4)
12. Melbourne -3.15 (+2)
13. Port Adelaide -3.23 (-4)
14. Adelaide -3.3 (-8)
15. Gold Coast -7.0
16. North Melbourne-10.8 (+1)
17. Sydney -11.6 (-1)
18. Carlton -18.2

It's so close in the middle that Adelaide dropped four points, and eight places. I know it's early days, but four standout teams so far.

Brisbane v Collingwood +10
North Melbourne v Essendon +16
West Coast +39 v Port Adelaide
GWS +26 v Fremantle
Melbourne v St Kilda +3
Richmond +16 v Sydney
Western Bulldogs +17 v Carlton
Adelaide +14 v Gold Coast
Hawthorn v Geelong +21

Essendon in 5th is closer to Gold coast in 15th than Collingwood in 4th

 

[tazaa]

When was the last time Sydney were on bottom of the Squiggle rankings?

Never
I thought you guys would struggle but then do a Geelong and keep hanging around and prove us wrong.
Seems time has finally caught up with the Swans. That final was the nail

 

[iluvparis]

They will be, always rising.

Didnt do analysis this week these were thrown together this arvo. Next week will post the updated rankings and a brief sentence as to why.

For the record my rankings aren't based on a data model, they are just my personal rankings of the teams from best to worst, so early season is heavily weighted to my preseason rankings of a side before I think its definitely proven otherwise. They're far from perfect and I know that, but it's a bit of fun.

Probably doesn't belong in an 'analytics' thread then?

 

[Final Siren]

Final Siren

I tipped 6 upsets with the tipping system you helped shape. I got 7/9 and won the round in my 95-person work comp to get the weekly prize. Now just one point off first too.

Thanks for your help. I’d say that the system is a minor success (made slightly more money than the entry fee) even after 4 rounds.

That was all you! I didn't do anything.

It's a good system and this has been the perfect year for it. If you'd tipped the underdog every time the favorite was 50-67% and the favorite when they were better than 67%, you'd have 7/9, 7/9, 2/9, and 7/9 so far, for 23 tips total.

The challenge is probably sticking to it even when the favorites have a good run and your fallible human brain starts suggesting tweaks.

 

[Suma Magic]

That was all you! I didn't do anything.

It's a good system and this has been the perfect year for it. If you'd tipped the underdog every time the favorite was 50-67% and the favorite when they were better than 67%, you'd have 7/9, 7/9, 2/9, and 7/9 so far, for 23 tips total.

The challenge is probably sticking to it even when the favorites have a good run and your fallible human brain starts suggesting tweaks.

Cheers.

The system is tip faves when:
- they have an 80%+ chance of winning according to the odds
- for 51-79% faves, tip faves if (tip % / chance of winning) is less than 10%.

Eg this week crows are currently 79% chance of winning. So don’t tip on the first condition. 84% of tipsters have backed them. So 84/79 is 7% diff. So as it’s less than 10% I tip the crows.

This has given me 21.

I think sticking to it should be easy, assuming that I drop off the overall lead is I guess I’m likely to.

 

[Chief]

Probably doesn't belong in an 'analytics' thread then?

I think it's interesting to see whether someone's personal assessment does or looks better than data models. Not my area though.

 

[Final Siren]

I think it's interesting to see whether someone's personal assessment does or looks better than data models. Not my area though.

I think there's enough evidence that human assessment is definitely worse than a model, unless the human gets lucky for a while or is quite exceptional. Even the experts, who spend all day applying their human brain to the task, usually perform worse, as evidenced by your local newspaper's tips.

But I'm sure a human assessment often LOOKS better! I've seen lots of this on BigFooty, where people post rankings that are approved at the time by the hive mind, but then when you look at them later they were wildly wrong -- and not just wrong in the way where everyone was wrong (like how no-one tipped the last three premiers before the start of the season), which is understandable, but in the way where everything they claimed to have special insight about turned out to be off. That's fun to read about, and attracts attention, but is kind of unforgivable for a forecaster, since you're taking common wisdom and making it worse.

On the flip side, there is also a kind of selective memory when people make big calls and get them right! At first, this attracts ridicule, but if it turns out to be correct, popular sentiment seems to switch around to, "Yeah, but was that really surprising?" Everything seems obvious in retrospect, once it's already happened. So people underestimate what a great call it was, and forget that not many others made it.

Anyway, I guess I'm saying that if you wanted to compare the accuracy of human & computer forecasters, you should do so with real evidence, not just try to remember who posted what when.