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

Remove this Banner Ad

One thing I dont get with the current forecast

Qualifying final - MCG - Western Bulldogs 79 - 78 Richmond
Grand Final - MCG - Richmond 79 - 77 Western Bulldogs

Is this because it views the QF as a Bulldogs home game while the GF is viewed as neutral?
Great question. It's because I revamped the Home Ground Advantage algorithm over the off-season. COVID provided some pretty good support for the importance of crowd support, so now Squiggle estimates the relative level of fan support for each team.

I haven't actually looked properly to see how much sense this makes in finals, but to explain the results you're seeing, the first game is a Bulldogs home game and the second is a Richmond home game. They're at the same venue, but still, more of the home team's fans turn up in home games. So Richmond get +1 point of HGA in the first and +3 points in the second.
 
Great question. It's because I revamped the Home Ground Advantage algorithm over the off-season. COVID provided some pretty good support for the importance of crowd support, so now Squiggle estimates the relative level of fan support for each team.

I haven't actually looked properly to see how much sense this makes in finals, but to explain the results you're seeing, the first game is a Bulldogs home game and the second is a Richmond home game. They're at the same venue, but still, more of the home team's fans turn up in home games. So Richmond get +1 point of HGA in the first and +3 points in the second.
Relatedly, just before COVID, I built a crowd predictor based on some amazing work from Tony Corke (Matter of Stats):


This provides a remarkably good estimate of how many people will turn up to each game (and why).

But since crowd numbers immediately plummeted to zero, and are still affected today, it only works as a kind of "what-if" simulator of how many people would have come along to the footy if there had been no pandemic.

For example, Collingwood v Brisbane at Docklands had 28,057. Under normal circumstances, it would have been about 37k:
Code:
$ ./estimate-crowd.pl 6258
Fetching data from Squiggle API...
[6258] 2021 Round 3 Collingwood v Brisbane Lions @ Docklands
            +0 Start                                         0
       +36,552 Intercept                                36,552
        -3,696 Period                                   32,856
        +4,171 Home team (Collingwood)                  37,027
          +990 Away team (Brisbane Lions)               38,017
       +16,073 Home team members (85k)                  54,090
          +201 Away team members (28k)                  54,291
        -9,367 Venue (Docklands)                        44,924
        -1,582 Day of week (Thursday)                   43,342
        -3,226 Time (19:40)                             40,116
        +2,242 Time + Day of Week                       42,358
        -5,791 Month (Apr)                              36,567
        -1,022 Home is favourite (1)                    35,545
        +1,573 Home favourite + team                    37,118
           -38 Favourite strength (2pts)                37,080
            +0 Vic rivals                               37,080
            +0 Same state (VIC/QLD)                     37,080
            +0 Top 8 (neither)                          37,080
            +0 Top 8 + Month                            37,080

Other games in Round 3:
  • NTH v BUL was 28k instead of 32k
  • ADE v GCS was 30k instead of 43k
  • RIC v SYD was 44k instead of 55k
  • ESS v STK was 29k instead of 40k
  • WCE v PRT was 42k instead of 49k
  • CAR v FRE was 25k instead of 29k
  • GWS v MEL was 10k, as expected
  • GEE v HAW was 50k instead of 63k
 

Log in to remove this ad.

Oh that's a bit weird. I've made a tweak.
I recall finding from your site something that rated teams vs other teams based on the weeks results (contested win, dishonourable loss) but now can’t find it.

And loved “providence”
 
Bonus Melbourne 2013 facts!

Melbourne 2013 were so bad, when they lost by 94 points in Round 3, their percentage improved (from 28 to 35).

Melbourne 2013 went 2-20 at 54.1%, but didn't finish bottom, because GWS only won 1 game all year.

Melbourne 2013's only wins came against GWS after a mighty come-from-behind effort in the last quarter, and a flop over the line by 3 points against the Bulldogs (who finished 15th).

The nadir for Melbourne was probably Round 3, 2014, after losing to GWS by five goals. This was the worst a team had been since St Kilda 1956, who similarly came off a near-winless year and opened the new season with three bad losses.

Here's an old pic of every team on Squiggle from 1995-2016. Guess which one is Melbourne 2013!
View attachment 1094537

Probably because the AFL didn't want to award GWS a priority pick, Melbourne 2013 didn't get one, either, despite being a historically terrible team in a priority pick era. Here is a chart of bad teams, with icons for those who got priority picks, and crosses for those who didn't. The really bad teams with crosses are mostly the two expansion teams and... guess who!

View attachment 1094538
I wasted 11 weekends on Melbourne in 2013. No wonder I have issues.
 
I wasted 11 weekends on Melbourne in 2013. No wonder I have issues.

I believe the Melbourne 2013 season review DVD is shown on a loop to terrorist suspects in Guantanamo Bay.
 
Those two Melbourne wins in 2013 were bloody weird
The first one, in R4, trailed GWS by 19pts at three quarter time then stacked on 12 goals in the last quarter to win by 41. 12 goals! They ended up only averaging 10 goals per game for the year, and kicked 12 in one quarter.
The other, in R14, they kicked a goal with 12 minutes left to go 44 points in front of the Bulldogs, who then kick 7 unanswered goals in 9 minutes of game time, and end up just getting held off in the last minute.
 
Great question. It's because I revamped the Home Ground Advantage algorithm over the off-season. COVID provided some pretty good support for the importance of crowd support, so now Squiggle estimates the relative level of fan support for each team.

I haven't actually looked properly to see how much sense this makes in finals, but to explain the results you're seeing, the first game is a Bulldogs home game and the second is a Richmond home game. They're at the same venue, but still, more of the home team's fans turn up in home games. So Richmond get +1 point of HGA in the first and +3 points in the second.
I guess grand finals don’t have “home” teams though.
 
Well, yes and no. The crowd will be more diverse, but some teams get to play 15 times per year at the MCG and some teams play there twice if they’re lucky.
Yes, but the post in question was about crowd support and not ground familiarity.
 

(Log in to remove this ad.)

Delusional if you think Victorian teams aren't significantly advantaged in crowd numbers against non Vic sides in grand finals. Absolutely delusional.
I didn’t say that. And we were talking about Western Bulldogs v Richmond.
 
Bonus Melbourne 2013 facts!

Melbourne 2013 were so bad, when they lost by 94 points in Round 3, their percentage improved (from 28 to 35).

Melbourne 2013 went 2-20 at 54.1%, but didn't finish bottom, because GWS only won 1 game all year.

Melbourne 2013's only wins came against GWS after a mighty come-from-behind effort in the last quarter, and a flop over the line by 3 points against the Bulldogs (who finished 15th).

The nadir for Melbourne was probably Round 3, 2014, after losing to GWS by five goals. This was the worst a team had been since St Kilda 1956, who similarly came off a near-winless year and opened the new season with three bad losses.

Here's an old pic of every team on Squiggle from 1995-2016. Guess which one is Melbourne 2013!


Probably because the AFL didn't want to award GWS a priority pick, Melbourne 2013 didn't get one, either, despite being a historically terrible team in a priority pick era. Here is a chart of bad teams, with icons for those who got priority picks, and crosses for those who didn't. The really bad teams with crosses are mostly the two expansion teams and... guess who!

View attachment 1094538

Hawks Eagles Tigers Magpies Bulldogs went on to flags GWS Gold coast Carlton did not . Saints and freo came 'close' -spot the pair

Since 2000 the first group have 12 flags of the 21 flags on offer. no wonder theres still fans of bottoming out -personally I think the era is passed
 
Those two Melbourne wins in 2013 were bloody weird
The first one, in R4, trailed GWS by 19pts at three quarter time then stacked on 12 goals in the last quarter to win by 41. 12 goals! They ended up only averaging 10 goals per game for the year, and kicked 12 in one quarter.
The other, in R14, they kicked a goal with 12 minutes left to go 44 points in front of the Bulldogs, who then kick 7 unanswered goals in 9 minutes of game time, and end up just getting held off in the last minute.
Love the Dees, what a fn team.
 
Hawks Eagles Tigers Magpies Bulldogs went on to flags GWS Gold coast Carlton did not . Saints and freo came 'close' -spot the pair

Since 2000 the first group have 12 flags of the 21 flags on offer. no wonder theres still fans of bottoming out -personally I think the era is passed
Richmond 1983-2012 destroyed my faith in the rebuild. Turns out you actually can be bad year after year, with no end in sight, if your club is run by clowns who throw away draft picks. (See also: Gold Coast.)
 
Oh, we're apparently making finals now according to the Squiggs instead of finishing 9th after a couple of wins. Lemme guess, if we beat Port we'll be minor premiers? :drunk:
squiggle can sniff my butthole

On SM-G925I using BigFooty.com mobile app
That utter blasphemy of the Squiggle is not tolerated here, sir.
 
That was a ripper weekend for Squiggle. Seven tips and much closer to the actual margins than the market.

The preseason ladder prediction is looking good, too. Missed Sydney, but so did everyone (except Damien Barrett), and it's getting killed by Geelong, but calls are looking good on the Bulldogs, Melbourne, Brisbane, Carlton, and West Coast.
 
That is interesting. 4 games in usually sorts things out and 17 teams still rated for the 8 (or second national draft pick up for the taking on a creditable 8 wins) depends how you look at it

View attachment 1100118
Imagine if that had been the top 6 in 2004

Bulldogs - 43 years since last granny
Melbourne - 40 years since last flag
Sydney - 71 years since last flag
Richmond - 2 finals series in last 21 seasons
Port - haven't made a granny
Geelong - 41 years since last flag

Not to mention the Bulldogs, Dees, Tiges and Cats all being rubbish the previous season. That finals series would go off!
 

Remove this Banner Ad

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

Remove this Banner Ad

Back
Top