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

[Holy Rioli]

Top 2 chances:
100% Richmond
41% West Coast
20% Collingwood
16% Port Adelaide
12% Melbourne
7% GWS
3% Geelong
2% Hawthorn

Hello Mr Final Siren.

Hawthorn currently have the 3rd best % in the AFL, 4th best attack, 3rd best defense. Why does squiggle rank them so low?

If we turned at 15 point lost to WCE in round 9 into a 1 point win, we'd actually be 2nd on the ladder right now.

 

[Gralin]

Hello Mr Final Siren.

Hawthorn currently have the 3rd best % in the AFL, 4th best attack, 3rd best defense. Why does squiggle rank them so low?

If we turned at 15 point lost to WCE in round 9 into a 1 point win, we'd actually be 2nd on the ladder right now.

it's chance to get top 2 from here, doesn't just require us to win games, requires teams above us to drop games, you have to look at the fixture

 

[Final Siren]

Hello Mr Final Siren.

Hawthorn currently have the 3rd best % in the AFL, 4th best attack, 3rd best defense. Why does squiggle rank them so low?

If we turned at 15 point lost to WCE in round 9 into a 1 point win, we'd actually be 2nd on the ladder right now.

Hawthorn is rated close to 2nd by Squiggle. But so are a lot of teams.

The main reason the Hawks aren't higher, I think, is they've had three shockers.

Good wins:

• R1 v Collingwood
• R2 v Geelong (only by 1 pt though)
• R4 v Melbourne
• R13 v Adelaide
• R16 v Bulldogs

Bad losses:

• R5 v North Melbourne
• R9 v Brisbane
• R17 v Brisbane

 

[DrEvil_]

Hello Mr Final Siren.

Hawthorn currently have the 3rd best % in the AFL, 4th best attack, 3rd best defense. Why does squiggle rank them so low?

If we turned at 15 point lost to WCE in round 9 into a 1 point win, we'd actually be 2nd on the ladder right now.

Because we are pretty shit mate, the best I can say is the Hawks are average. Take off your Hawthorn glasses.

 

[Hobbes]

Essendon +22 v Sydney
Richmond +9 v Collingwood
Geelong +3 v Brisbane
GWS +35 v St Kilda
Gold Coast +21 v Carlton
Adelaide v Melbourne +4
North Melbourne v West Coast +16
Western Bulldogs v Port Adelaide +23
Fremantle v Hawthorn +17

7/9. If the algorithm knew about North's record in Tasmania, it might have done better still.

1. Richmond 27.2
2. GWS 18.1
3. Hawthorn 14.4 (+4)
4. Essendon 13.1 (+1)
5. Collingwood 11.6 (-1)
6. Melbourne 11.4
7. Geelong 8.5 (+3)
8. West Coast 8.4 (-5)
9. Port Adelaide 7.0
10. Brisbane 2.2 (-2)
11. Adelaide -0.9
12. North Melbourne -2.4 (+1)
13. Sydney -6.5 (-1)
14. St Kilda -7.7
15. Fremantle -21.3
16. Western Bulldogs -28.6
17. Gold Coast -31.9
18. Carlton -32.0

Richmond +19 v Geelong
Hawthorn +1 v Essendon
Brisbane +11 v North Melbourne
Adelaide v Port Adelaide +8
St Kilda +21 v Western Bulldogs
Sydney v Collingwood +13
Carlton v GWS +45
Melbourne +49 v Gold Coast
West Coast +30 v Fremantle

And the predictive ladder

1. Richmond 17.4
2. West Coast 15.2
3. Collingwood 14.6
4. GWS 14.5
5. Port Adelaide 14.1
6. Melbourne 13.64 (+1)
7. Geelong 13.60 (+1)
8. Hawthorn 13.45 (-2)
9. Sydney 12.2
10. North Melbourne 12.1 (+1)
10. Essendon 12.0 (-1)
12. Adelaide 11.0 (-1)
13. Fremantle 8.3
14. Brisbane 6.4
15. St Kilda 6.2 (+1)
16. Western Bulldogs 6.0 (-1)
17. Gold Coast 4.2
18. Carlton 3.1

With the end of the year approaching, a quick look at the ladder if every favourite (per the algorithm) wins.

Richmond 18
Collingwood 16
GWS 15.5
Hawthorn 15 121.2
WCE 15 118
Port Adelaide 14
Melbourne 13 125
Geelong 13 120
North Melbourne 12 109
Essendon 12 103
Sydney 11 103
Adelaide 11 99
Fremantle 8
Brisbane 7
Western Bulldogs 6
St Kilda 5.5
Gold Coast 4
Carlton 2

13 wins looks like enough, although Adelaide's goose is probably cooked even if they win their last four - their dire percentage will probably kill their hopes. For the top 4, 15 might be enough, or not.

 

[Holy Rioli]

Because we are pretty shit mate, the best I can say is the Hawks are average. Take off your Hawthorn glasses.

Please tell me how I'm wearing glasses mate? I quoted facts. Fact is we have the third best %, 4th best attack, 3rd best defense but rated 7th by Squiggle. I was just simply asking why.

 

[Holy Rioli]

it's chance to get top 2 from here, doesn't just require us to win games, requires teams above us to drop games, you have to look at the fixture

I know that (if we had beat WCE earlier we'd be 2nd right now). I just found it strange that squiggle rated us 7th (and flagpole has us missing finals) when we have the 3rd best % in the AFL.

I was thinking it was down to 2 things - our losses to the Lions, and the fact we are pretty accurate in front of goal, and squiggle adjust scores based on scoring shots instead of actual score.

 

[spinynorman]

I know that (if we had beat WCE earlier we'd be 2nd right now). I just found it strange that squiggle rated us 7th (and flagpole has us missing finals) when we have the 3rd best % in the AFL.

You’re also 7th on the ladder so it isn’t exactly wildly rating you different to where you’re sitting.

 

[moond0g]

Please tell me how I'm wearing glasses mate? I quoted facts. Fact is we have the third best %, 4th best attack, 3rd best defense but rated 7th by Squiggle. I was just simply asking why.

When you say you have the 4th best attack and 3rd best defense you're just taking the for and against scores. These scores calculate the %. So you've basically just used the same stats twice

 

[iluvparis]

When you say you have the 4th best attack and 3rd best defense you're just taking the for and against scores. These scores calculate the %. So you've basically just used the same stats twice

But mate - i believe they are also rated quite highly on points differential!!!

 

[Holy Rioli]

When you say you have the 4th best attack and 3rd best defense you're just taking the for and against scores. These scores calculate the %. So you've basically just used the same stats twice

But mate - i believe they are also rated quite highly on points differential!!!

Actually I haven't, it is very possible for a team to have the 3rd best defense, 4th best attack and not have the 3rd best %. Melbourne has the best attack and 10th best defense but second best %.

Melbourne 1868 points for (1st) 1467 against (10th) % = 127.3 (2nd)
Richmond 1754 points for (2nd) 1273 against (2nd) % = 137.8 (1st)
Collingwood 1680 points for (3rd best) 1422 against (9th) % = 118.1
Hawthorn 1631 points for (4th best) 1329 against (3rd) % = 122.7

Maybe take the time to learn some basic maths before shooting your mouth off.

 

[living_in_syd]

Top 2 chances:
100% Richmond
41% West Coast
20% Collingwood
16% Port Adelaide
12% Melbourne
7% GWS
3% Geelong
2% Hawthorn

How are richmond 100%?

We are only two games clear from 4th and can lose all 4 remaining games

 

[Final Siren]

How are richmond 100%?

We are only two games clear from 4th and can lose all 4 remaining games

Rounding. It's 99.57% if you go to decimals.

Put another way, Richmond make top 2 in 29,871 out of 30,000 simulations.

Anything can happen in football, but a particular series of simultaneous upsets are required to push the Tigers out of top 2 now -- not only do the Tigers have to lose matches they should win, but other teams also have to win matches they should lose.

Chains of probability like that shrink the odds dramatically -- there is a 10% chance of a 1-in-10 event happening, but only a 1% chance of two 1-in-10 events happening together.

 

[Billy ray]

Rounding. It's 99.57% if you go to decimals.

Put another way, Richmond make top 2 in 29,871 out of 30,000 simulations.

Anything can happen in football, but a particular series of simultaneous upsets are required to push the Tigers out of top 2 now -- not only do the Tigers have to lose matches they should win, but other teams also have to win matches they should lose.

Chains of probability like that shrink the odds dramatically -- there is a 10% chance of a 1-in-10 event happening, but only a 1% chance of two 1-in-10 events happening together.

wooo! maths

so. would the suns win their remaining 4 games in 1 out of 30,000 simulations ? rounded up?

 

[bigpapagman]

wooo! maths

so. would the suns win their remaining 4 games in 1 out of 30,000 simulations ? rounded up?

No.

 

[Final Siren]

I know that (if we had beat WCE earlier we'd be 2nd right now). I just found it strange that squiggle rated us 7th (and flagpole has us missing finals) when we have the 3rd best % in the AFL.

I was thinking it was down to 2 things - our losses to the Lions, and the fact we are pretty accurate in front of goal, and squiggle adjust scores based on scoring shots instead of actual score.

Squiggle Flagpole actually downgrades the Hawks because they are in danger of missing finals.

For example, on form, Essendon are rated quite a bit above Sydney, but it's the other way around on Flagpole because the Bombers are unlikely to make Top 8 while Sydney still have a shot.

1. Normal squiggle -- team form rating.

2. Tower -- likely finishing positions.

3. Flagpole -- #1 modified by #2.

 

[Final Siren]

wooo! maths

so. would the suns win their remaining 4 games in 1 out of 30,000 simulations ? rounded up?

I just ran 100,000 sims and the best the Suns could manage was 3 out of 4. Which is still pretty great, considering they face Melbourne, Richmond, Brisbane, and Geelong.

 

[Billy ray]

I just ran 100,000 sims and the best the Suns could manage was 3 out of 4. Which is still pretty great, considering they face Melbourne, Richmond, Brisbane, and Geelong.

woo hoo!
damn I wish my team was good.... like at round 14 2014 on the squiggle....

 

[Gralin]

I just ran 100,000 sims and the best the Suns could manage was 3 out of 4. Which is still pretty great, considering they face Melbourne, Richmond, Brisbane, and Geelong.

Imagine if they actually beat Melbourne, Richmond and Geelong on the run home, i mean they beat the Swans and then lost to Carlton so its possible.....

 

[Billy ray]

Imagine if they actually beat Melbourne, Richmond and Geelong on the run home, i mean they beat the Swans and then lost to Carlton so its possible.....

You wouldn’t stop hearing about it from me that’s for sure

 

[Gralin]

You wouldn’t stop hearing about it from me that’s for sure

Thats what ignore is for

 

[NoobPie]

Rounding. It's 99.57% if you go to decimals.

Put another way, Richmond make top 2 in 29,871 out of 30,000 simulations.

Anything can happen in football, but a particular series of simultaneous upsets are required to push the Tigers out of top 2 now -- not only do the Tigers have to lose matches they should win, but other teams also have to win matches they should lose.

Chains of probability like that shrink the odds dramatically -- there is a 10% chance of a 1-in-10 event happening, but only a 1% chance of two 1-in-10 events happening together.

But in reality they are not mutually exclusive events

 

[iluvparis]

But in reality they are not mutually exclusive events

I'd say two games played between two different sets of teams are close enough to be mutually exclusive not to worry about it

 

[NoobPie]

I'd say two games played between two different sets of teams are close enough to be mutually exclusive not to worry about it

But multiple games played by the same team aren’t

 

[Final Siren]

But multiple games played by the same team aren’t

That's very true! Squiggle long-term predictions assume team strength will remain the same over multiple future rounds. But in reality, a team will often gain or lose form due to injuries or whatever. A team that beats expectations this weekend is more likely to beat (today's) expectations the weekend after as well.

So they are indeed not independent events.

This doesn't matter for regular squiggle game tips, since the squiggle ratings update. But that's not the case in future simulations of multiple rounds. Which means they are over-confident.