Swanks McSwankserton
Senior List
So over the covid hiatus I was going through one of the innumerous 'my team has more flags than you threads' and came across this wonderful post (thanks btw both 4 this & the inspiration BF Tiger ):
Data from:https://en.wikipedia.org/wiki/List_of_VFL/AFL_premiers
Apologies to most clubs for the colouring scheme but I couldn't figure out how to do alternating colours and once u decide the background is white that means collingwood has to be black and then essendon and st kilda are red and richmodn yellow and then you discover a surprising amount of clubs all have some shades of blue or yellow or red (looking at you adelaide) and u question why you still care anymore.
Interesting takeaways:
Collingwood has played in a rediculous amount of gfs. Like as many gfs as a third of the rest of the compertition combined (albeit a third of the comp is really new), and like almost a third of all gfs.
One successful stint really boosts you up the graph.
Anyways as great and cool and interesting as this is I wasn't really satisfied and reflected the concept of a "premiership window" and though it was all about prelims. A bunch of prelims r stupidly close and if u played in a prelim I reckon u wouldnt have needed an unreasonable amount more to go right to have won a flag, and also u probably had a really great year in which u were a genuine contender. So I graphed those:
Data from: https://afltables.com/afl/seas/[pick a year].html
It turns out pre early 90s they only played one prelim/year, so I just said a grand final appearance was a prelim win in those years. Unfortunately no-one collated all the prelim data nicely like on wikipedia so now ur clubs history began with their first prelim appearance. And I gave the bears prelim to the lions and didn't bother with too much care for names anymore cos it was inflating the legend. No more mister nice graph.
Anyways, this graph was super satisfying and by far the most interesting to sudy in detail. I thought there was finally enough data that a little statistics analyasis wouldn't be totally useless so as u saw I started that. I wanted to know what said stats said about average time between contention for clubs and average duration/longevity of contention for clubs.
Also in the vein of trying to capture a premiership window I decided to draw a line connecting all prelims withn two years of each other (naively I originally had it at 4, but then collingwood had like a 30 yr premiership window so I had to change it). It actually turned out really nicely and I'm interested in what people think about it, would port fans say the '04 & '07 gfs are part of the same era/premiership window (I wasnt following footy closely then)?, what about '15 and '18 eagles? '13 and '16 cats? I tentatively refer to these groupings as windows.
The stats next to each club: # of prelims, % of those prelimes won (yayyy, richmond!), # of windows, mean length of windows (yrs), mean length of yrs between prelims, mean length of yrs between windows. These statistics include time between last prelim and this season.
The stats at top of graph take into account all clubs: % of prelims that are a part of windows, mean windows length, mean length of prelim droughts, mean length of window droughts.
But then I was looking at those rediculous super runs of gf appearances (richmond 20s and essendon 40s, who knew) ending w/ 80s hawthorn, and thought they had to be throwing off the data for the modern day, and especially when u take into account salary cap, draft, 7 (i think) whole new clubs introduced, changes to finals structures, prelims mean something different nowadays and I quickly lost all faith in those statistics. So I reran my program starting where I knew the haters wanted it, 1990:
And to my surprise avg length of windows dropped less than a year, and time between windows dropped a year as well! Guess I forgot about those crazy super droughts that must've been counterbalancing the stats. The % didn't fall much either, so congrats to port and brisbane fans, there's a 3/4 chance each you'll make another prelim in the next 2 years. Whether that turns out to be more of a north '14-'15 kinda deal or a north '94-'99 stretch only time will tell.
So, in conclusion, super sorry for the essay this post turned out to be. Hope you guys found it as interesting as I did and learnt a bunch like I did too. Interested to hear peoples thoughts and takeaways on all this, especially the prelim == premiership window hypothesis. Was gonna share this earlier but thought I'd wait till this seasons gfs and prelims were done.
p.s. not sure why the '90 graph went on a exta yr, but the stats would be taking that into account unfortunately.
And I was looking at it and thinking about how little of the story of each year a graph can capture (and no graph will could ever capture the 100s and 1000s of characters and plotlines and just little moments that even a single season holds), and you can also sort of see the different successful eras a club has and I wondered how adding grand finals losses would shape them and if we could see a proper nice short history of the vfl/afl. So off I went cos covid was terribly boring and made a graph about gfs:
Data from:https://en.wikipedia.org/wiki/List_of_VFL/AFL_premiers
Apologies to most clubs for the colouring scheme but I couldn't figure out how to do alternating colours and once u decide the background is white that means collingwood has to be black and then essendon and st kilda are red and richmodn yellow and then you discover a surprising amount of clubs all have some shades of blue or yellow or red (looking at you adelaide) and u question why you still care anymore.
Interesting takeaways:
Collingwood has played in a rediculous amount of gfs. Like as many gfs as a third of the rest of the compertition combined (albeit a third of the comp is really new), and like almost a third of all gfs.
One successful stint really boosts you up the graph.
Anyways as great and cool and interesting as this is I wasn't really satisfied and reflected the concept of a "premiership window" and though it was all about prelims. A bunch of prelims r stupidly close and if u played in a prelim I reckon u wouldnt have needed an unreasonable amount more to go right to have won a flag, and also u probably had a really great year in which u were a genuine contender. So I graphed those:
Data from: https://afltables.com/afl/seas/[pick a year].html
It turns out pre early 90s they only played one prelim/year, so I just said a grand final appearance was a prelim win in those years. Unfortunately no-one collated all the prelim data nicely like on wikipedia so now ur clubs history began with their first prelim appearance. And I gave the bears prelim to the lions and didn't bother with too much care for names anymore cos it was inflating the legend. No more mister nice graph.
Anyways, this graph was super satisfying and by far the most interesting to sudy in detail. I thought there was finally enough data that a little statistics analyasis wouldn't be totally useless so as u saw I started that. I wanted to know what said stats said about average time between contention for clubs and average duration/longevity of contention for clubs.
Also in the vein of trying to capture a premiership window I decided to draw a line connecting all prelims withn two years of each other (naively I originally had it at 4, but then collingwood had like a 30 yr premiership window so I had to change it). It actually turned out really nicely and I'm interested in what people think about it, would port fans say the '04 & '07 gfs are part of the same era/premiership window (I wasnt following footy closely then)?, what about '15 and '18 eagles? '13 and '16 cats? I tentatively refer to these groupings as windows.
The stats next to each club: # of prelims, % of those prelimes won (yayyy, richmond!), # of windows, mean length of windows (yrs), mean length of yrs between prelims, mean length of yrs between windows. These statistics include time between last prelim and this season.
The stats at top of graph take into account all clubs: % of prelims that are a part of windows, mean windows length, mean length of prelim droughts, mean length of window droughts.
But then I was looking at those rediculous super runs of gf appearances (richmond 20s and essendon 40s, who knew) ending w/ 80s hawthorn, and thought they had to be throwing off the data for the modern day, and especially when u take into account salary cap, draft, 7 (i think) whole new clubs introduced, changes to finals structures, prelims mean something different nowadays and I quickly lost all faith in those statistics. So I reran my program starting where I knew the haters wanted it, 1990:
And to my surprise avg length of windows dropped less than a year, and time between windows dropped a year as well! Guess I forgot about those crazy super droughts that must've been counterbalancing the stats. The % didn't fall much either, so congrats to port and brisbane fans, there's a 3/4 chance each you'll make another prelim in the next 2 years. Whether that turns out to be more of a north '14-'15 kinda deal or a north '94-'99 stretch only time will tell.
So, in conclusion, super sorry for the essay this post turned out to be. Hope you guys found it as interesting as I did and learnt a bunch like I did too. Interested to hear peoples thoughts and takeaways on all this, especially the prelim == premiership window hypothesis. Was gonna share this earlier but thought I'd wait till this seasons gfs and prelims were done.
p.s. not sure why the '90 graph went on a exta yr, but the stats would be taking that into account unfortunately.
