Certified Legendary Thread The Squiggle is back in 2023 (and other analytics)
- Thread starter Final Siren
- Start date
- Tagged users None
Forward Press
Hall of Famer
It was only 27 points, and you only had 2 scoring shots more than Brisbane.
It was also a notional home game, but on this point I am not sure how Final Siren adjusts it.
I didn't see the original tip but given Brisbane are a slightly below average defensive team on the squiggle, Geelong's predicted score would've been pretty close to what they did score, especially when it is adjusted for Geelong's accurate kicking.
The big surprise in that game was Geelong restricting Brisbane, the highest rated attacking team on the squiggle, to just 46 points, hence the noticeable shift right.
- Aug 18, 2009
- 4,229
- 17,495
- AFL Club
- Richmond
- Thread starter
- #1,904
Good explanations above; also, sideways movement is a good thing, as long as it's rightwards. That means your defensive rating is increasing. Upwards means your attack rating is increasing.
Home advantage wasn't a factor in this match: less than 1 point in Geelong's favour. Geelong received a mild benefit from In/Outs, since those were expected to favour Brisbane.
- Aug 18, 2009
- 4,229
- 17,495
- AFL Club
- Richmond
- Thread starter
- #1,905
It's pretty rare! Since 2011, which is the dataset I happen to have on hand, and excluding finals rounds, it's only happened twice:
- 2016 Round 14 (6 games)
- 2016 Round 1 (9 games)
- 2019 R9
- 2019 R7
- 2017 R23
- 2017 R18
- 2015 R23
- 2015 R15
- 2014 R15
- 2012 R8
- 2012 R2
So it's pretty crazy that it happened last round despite no home teams being at home.
With regard to gathering adequate data, it depends on what you're trying to establish. We have a lot of match data; things only become dicey when you're trying to test a theory that applies to a small slice of it. That's why I always try to establish a hypothesis that applies equally to every team in the competition, e.g. "teams perform better when they have more crowd support" rather than "Hawthorn play better at the SCG" or "Richmond have Brisbane's measure at the Gabba." The latter kinds of theories are basically impossible to verify statistically, because they rely on maybe half a dozen match results - or else become inherently nonsensical, like what does Richmond 2005 really have in common with Richmond 2019? Not players, that's for sure.
I'm also wary of that idea that this year is unlike any other. It is a strange and chaotic year, of course, but people always find reasons why this year/game/team is unlike any other, and therefore shouldn't follow historical trends, so anything could happen. In my opinion, the reality is that every game/round/team/year is unique, but in relatively small ways, which don't exempt anyone from what we know about how football generally works. So, yes, outliers happen, but true ones are vanishingly rare, in my opinion, and I basically never exclude them. Instead, I think a good theory (like "crowds affect HGA") should broadly apply to everyone in all circumstances - it doesn't have to be perfectly accurate all the time, but it should be more right than wrong on average across all games.
- Aug 18, 2009
- 4,229
- 17,495
- AFL Club
- Richmond
- Thread starter
- #1,906
By the way, Squiggle's Who Won the Round is mostly functional nowadays, rating the importance of results. So for example, Geelong are #1 from last round for how their win over Brisbane raised their Top 4 chances. Biggest loser from last round was the Bulldogs, dropping from ~65% of Top 8 to ~40%.
Last edited:
Forward Press
Hall of Famer
That's really quite remarkable that the last time all 9 home teams won was over 4 years ago - and that this was a round where no team played in their home State, much less ground.
- Aug 18, 2009
- 4,229
- 17,495
- AFL Club
- Richmond
- Thread starter
- #1,908
Yep!
It could be a while before we see the next one, too. Home teams have won 56.9% of games since the 18-team comp began in 2012. If we get 20 non-bye, non-finals rounds of 9 games per year, we should see one where all home sides win about once every 8 years. (56.9%**9 = 0.62%, or once every 160 rounds.)
There has never been a 9-game round where all the away sides won*. We should expect one of those every 97 years. (43.1% ** 9 = 0.05%, or once every 1,944 rounds.)
(* We went close once: in R16 2012, away teams won every game except the first one, the streak ruined by Carlton failing to get over North. Richmond, playing a nominal home game in Cairns, managed to lose to a 0-14 Gold Coast side that weekend, but no, Carlton still had to mess things up. Also in R11 the same year, away teams won all matches of a 6-game bye round.)
iluvparis
Import Whisperer
- Apr 1, 2005
- 50,055
- 40,651
- AFL Club
- Carlton
- Other Teams
- Calgary Flames, Man Utd
The facts were in the article you just wrote off with no contradictory evidence. Ball constantly in dispute? Seems a stretch. And the US/Europe argument is irrelevant because FS produced an article in the Australian context.
- Feb 25, 2013
- 51,222
- 58,286
- AFL Club
- Brisbane Lions
I genuinely don't think you've been to an AFL game before. The difference in noise when one team has the ball in play compared to the other team is negligible at best. Even when a player has the ball or the opportunity to impact the game, the time taken in doing so is so minimal that to suggest the crowd impacts their ability to do so is ludicrous. When they're lining up for a set shot is really the only time it would have an impact.
The article posted seems to have used average HGA, not average HGA against teams traveling to Sydney, so really it's not exact. Not to mention rugby is a completely different game to football, one team has the possession before it is turned over, similar to Basketball or American Football where the crowd has a much greater opportunity to impact the play
iluvparis
Import Whisperer
- Apr 1, 2005
- 50,055
- 40,651
- AFL Club
- Carlton
- Other Teams
- Calgary Flames, Man Utd
Produce some numbers to support your hypothesis and to counter theirs and I will consider it.
- Feb 25, 2013
- 51,222
- 58,286
- AFL Club
- Brisbane Lions
Since 2000 home teams playing against teams from the same state win 52% of the time
When playing teams from interstate however, they win over 60%
To add to this, in the past when many Victorian teams had their own home grounds, the win % against other Victorian teams were higher
Last edited:
iluvparis
Import Whisperer
- Apr 1, 2005
- 50,055
- 40,651
- AFL Club
- Carlton
- Other Teams
- Calgary Flames, Man Utd
Probably because the crowd differentiation is greater when they play an interestate side. Thanks for confirming the hypothesis of FS
- Feb 25, 2013
- 51,222
- 58,286
- AFL Club
- Brisbane Lions
It's the exact same scenario as the article previously quoted actually, just using more accurate data and relating it to the actual sport we're talking aboutProbably because the crowd differentiation is greater when they play an interestate side. Thanks for confirming the hypothesis of FS
The numbers show that crowd advantage is a small factor when compared to travel, familiarity, etc.
As traveling conditions have improved, HGA has decreased, with win % dropping from 64% in the 90s down to 58% this decade
Similarly, as Victorian teams have been confined mostly to the MCG and Telstra Dome as opposed to their own suburban grounds, Vic v Vic HGA has also decreased, from 56.3% in the 80s to 52% currently
iluvparis
Import Whisperer
- Apr 1, 2005
- 50,055
- 40,651
- AFL Club
- Carlton
- Other Teams
- Calgary Flames, Man Utd
You realise the exact same numbers could be a result of crowd differential though right? \
- Why was HGA higher at suburban grounds?
- Why is HGA bigger against an interestate side?
You are literally producing evidence FOR the FS hypothesis - not against it.
- Feb 25, 2013
- 51,222
- 58,286
- AFL Club
- Brisbane Lions
Because of familiarity
Because of travel and familiarity
If you want this to be the case you would have to prove crowd differentials in suburban grounds, as well as in interstate games
GROTTO
TheBrownDog
- Jul 5, 2013
- 66,745
- 89,556
- AFL Club
- Adelaide
- Other Teams
- ¯\_(ツ)_/¯
Im a little surprised Carlton, Port and Freo wins were not all ranked in the top 4.
I can appreciate you have Geelong ranked 1 for the reasons you outlined. Not sure I would rank Pies over Hawks and Dons over North as better wins compared to Blues, Port and Freo. If the Bulldogs were the biggest loser, wouldnt the Blues win make them one of the biggest winners of the round and place them in the top 3 for the round?
iluvparis
Import Whisperer
- Apr 1, 2005
- 50,055
- 40,651
- AFL Club
- Carlton
- Other Teams
- Calgary Flames, Man Utd
No - it is because of crowd differential - as highlighted by the article posted by FS and the data you yourself posted which you have been unable to interpret correctly. Unless you are going to argue crowds were 50:50 at the old suburban grounds which would be a laughable claim.
- Aug 18, 2009
- 4,229
- 17,495
- AFL Club
- Richmond
- Thread starter
- #1,919
That looks like a chart from Matt Cowgill? He has written in favour of crowd-generated HGA in the AFL.
- Aug 18, 2009
- 4,229
- 17,495
- AFL Club
- Richmond
- Thread starter
- #1,920
Geelong's win was especially fine because not only did it give the Cats a boost, but it knee-capped a rival for a Top 4 or even Top 2 spot (Brisbane). So it was something like your classic "8 point game."
That wasn't really the case for Bulldogs v Carlton, who aren't clearly competing for limited spots in the same way.
Additionally, another potential Top 4 rival for the Cats in Hawthorn had a bad weekend. (Although Collingwood had a good one.) Rankings are affected by the results of other teams, especially when those teams are more clearly competing for the same spots.
Also there's also some special sauce in that the tool considers it especially meaningful that the Bulldogs' loss shifted them from "probably will make finals" (63%) to "probably won't" (41%). Carlton's win was excellent but it also only raised their projected finish from 13th to 11th, which is just not that big a deal.
GROTTO
TheBrownDog
- Jul 5, 2013
- 66,745
- 89,556
- AFL Club
- Adelaide
- Other Teams
- ¯\_(ツ)_/¯
With the likelihood of the Finals series played in QLD will this be a major variable in the Squiggle analysis of flag chances?
Are the Lions in the box seat? Does the chances of the Suns to make Finals dramatically improve?
If I was a journo, I'd use this as the basis for a weekly article. Real good stuff.
- Feb 25, 2013
- 51,222
- 58,286
- AFL Club
- Brisbane Lions
My article actually relates to the sport we are talking about, unless you wish to talk about the NRL in which I couldn't care less
- Feb 25, 2013
- 51,222
- 58,286
- AFL Club
- Brisbane Lions
No idea who did it, I saw an article in favour of HGA, which I'm sure crowd is a part of, never said it wasn't, just that it's a smaller part of it
- Aug 18, 2009
- 4,229
- 17,495
- AFL Club
- Richmond
- Thread starter
- #1,925
If the whole season winds up being played in hubs, then yeah, it'll almost certainly be the most unbalanced fixture of all time in terms of HGA - even with heavy discounting for the lack of crowds, and despite the AFL's best efforts to even things out (like playing Geelong v Brisbane at the SCG).
I don't think it's worth getting worked up about yet, because there's still so much uncertainty, and HGA is far less important than form (or goalkicking accuracy!). For example, Gold Coast had a decent net HGA last year and didn't win a game after Round 4. You still actually have to be good.
For finals, the AFL may do a similar thing whereby a Cats v Lions final would be held in Sydney rather than at the Gabba. Even if they don't, the Lions might not benefit anyway, e.g. they might earn home finals, or miss the 8 entirely.
So while it would be very handy, obviously, to wind up with an unearned home final or two, from where we stand right now, I don't think it moves the needle that much. (Not that I, a Richmond fan, would know anything about undeserved home finals.)
