Swanks McSwankserton
Senior List
So, I've been reading bigfooty for a bit now and noticed there are a lot of threads where debate over interstate vs victorian advantages and disadvantages is brought up, but long as they went on for no-one ever brought up any stats. I got really bored in the off season waiting for footy and uni to start up again and really enjoyed doing a similar project on graphing prelims last year during the boring as footy corona break so I decided to check out the results of interstate v vic games, and I thought I'd share it for those who might also be interested in such things. Without further boring ado...
Interstate v vic home and away games since South moved north, with draws counted as half a win:
Takeaways:
The ~48% of wins thing has been fairly consistent, with the only years interstate sides were in front on the all time count being '06 and '07. If you weight each year equally (rather than by total results like the graph does) you get an average win percentage of 47.7478%.
Interesting how the 2002-2007 interstate supremacy a. happened and b. was immediately followed by a 2008-2013 victorian supremacy. Not sure what this means (if anything? you flip enough coins you're bound to get unlikely streaks?), or if it will happen again?
(Subnote on balance of home games if you're interested: The AFL has been keeping it tight at exactly 50-50 with 3 exceptions: 1987 when they scheduled 1 extra interstate home game, 2020 with "4" extra interstate home games and in 2015, when for some random reason I can only logically assume was as an easter egg for nerds like me who were checking, for the first and only ever time the AFL scheduled an odd number of vic v interstate games, and rounded in favour of 1 extra victorian home game. It's too small to make out properly on the above graph but interstate sides ended up half a game in front of 50% that year too (same in 2020), with no exact 50% wins year season so far. As to what the actaul venues of these home and away games were I can't attest.)
Anyways I know what you're thinking, crazy interesting as that was, what was the actual result of all those games? How did those extra wins/losses translate onto that year's ladder? That brings us to the next graph...
You'll notice I graphed the 'expected' ladder positions, these values are based on the idea that in an 18 team competition that is "completely even" a group of say, 8 teams, selected "at random" would be expected to between them place 1st 8 times every 18 years. If you tell a computer to work backwards for you on those changing numbers, this is what you end up with.
This was one of the more surprising results of the whole thing; positions 1-4 and 6-7 are bang on their expected values. The sum of frequencies of actual position placemet is only ~0.04 in front on the sum of expected frequencies. If we have a look at the differences between expected and actual position placement for non finals positions (6: -1.274, 7: 0.926, 8: -1.074 (6-8 only takes into account the years they weren't finals positions), 9: -6.627, 10: 2.373, 11: 1.373, 12: 0.373, 13: 2.789, 14: -0.211, 15: 0.647, 16: -2.287, 17: 1.588, 18: 1.0) it turns out the impact of that extra home and away loss was felt by the poorer interstate sides in a year, and caused interstate sides as a group to miss out on finishing 9th rather than miss out on qualifying for finals.
Interstate sides are leaning a tad more 5th than 8th, but I don't read much into that. (Since the introduction of the current finals system, those remain the only two positions to never make the grand final, and in their annual showdown 5th is only 1 win in front of 8th! The record of 11-10 defies logic (imo 5th seems to be usually not too far off the top 4 and 8th usually just scapes/scraps into finals, so idk how this is the case)).
I could have just looked at the end total differences for finals sides as well instead of graph them, but I wanted to see if they were advanving at the expected rate or if the overwhelming majority of the progress happenned in a one off period. Most of them were advancing at the expected rate, especially 5th and 6th, but you can clearly see that between 2001-2007 (~25% of the total time (technically more like 18-19% but I added a bit, cos you couldn't expect many placements in early years; the per year number of interstate sides expected to be qualifying for finals didn't pass 2 till '94 (currently at 3.6 btw))) 50% of the total progress of 2nd, ~46% of 3rd, ~38% of 1st and 36% of 8th occurred.
Anyways, so interstate sides are qualifying for finals alright, how are they doing? How has a game being a final affected the win rate historically?
...positively it turns out, as funnily enough this is virtually bang on the expected value (50%) too.
But this doesn't show/capture which finals are being won. How much of that 50% are the big important finals and how much are the cheapy dud ones?
As you can see from the slight curves to the lines, over the years as more sides are added to the comp expected interstate side premierships have been accelerating as a group, while slowing down individually. The expected group premierships is currently going up ~0.45 premierships a year (which would be ~0.9 prelim wins (grand final appearances) a year) (and individual premierships going up ~0.055 a year), so at the end of this season we'd need an all interstate grand final to make the cosmic ledger balanced again for premierships and grand final appearances. (West coast expected premierships same as Brisbane's by the way, I just can't figure out how to alternate lines so they get drawn on top of each other, apologies (also apologies for grouping the bears and post merger lions together, but I felt they fulfilled the same function in this context)).
Anyways interstate sides have never been 2 premierships below nor 4 premierships above the expected line, and have spent 21 total seasons below the line and 18 above.
I feel I should point out here that I actually am heavily against the unfair scheduling of grand finals and prelims at the 'g, it is crazy rediculous. Worth pointing out here that this graph only has 12 data points to compare and work with, compared to the one and a half thousand in the first graph, which is my guess for why interstate sides aren't further behind than they are. I'm sure the FIXturing will definitely show up in the stats before (if) the MCG grand final contract ever expires...
Also worth noting that although the group is on track so far, individually it's more whacky. A year is a crazy long time, and with 18 teams each side is expected to lose a grand final, win 1 premiership than go into an immediate 16 year grand final draught before repeating again. So while Adelaide have a 22 year premiership drought so far (second longest of an interstate side after Freo), they're still about 3 years ahead of their expected premiership schedule, which means they've only just passed the halfway mark of their expected premiership drought (assuming the AFL doesn't add any more teams in the next 20 years, which would extend it). Meanwhile of all things Gold Coast is considered the second most hard done by interstate team on expected premierships (after Freo). Port then are the least hard done by of the sides in the -ve, depite having a premiership drought 7 years older than Gold Coast has been a thing.
All data used in graphs from (and much thanks to) afltables.com, this whole thing ended up being a much more fun and a more fantastic waste of time than I had hoped, learned a bunch, and I hope you found this (or at least some of it) as interesting as I did. Also apologies for the essay this turned out to be .
Interstate v vic home and away games since South moved north, with draws counted as half a win:
Takeaways:
The ~48% of wins thing has been fairly consistent, with the only years interstate sides were in front on the all time count being '06 and '07. If you weight each year equally (rather than by total results like the graph does) you get an average win percentage of 47.7478%.
Interesting how the 2002-2007 interstate supremacy a. happened and b. was immediately followed by a 2008-2013 victorian supremacy. Not sure what this means (if anything? you flip enough coins you're bound to get unlikely streaks?), or if it will happen again?
(Subnote on balance of home games if you're interested: The AFL has been keeping it tight at exactly 50-50 with 3 exceptions: 1987 when they scheduled 1 extra interstate home game, 2020 with "4" extra interstate home games and in 2015, when for some random reason I can only logically assume was as an easter egg for nerds like me who were checking, for the first and only ever time the AFL scheduled an odd number of vic v interstate games, and rounded in favour of 1 extra victorian home game. It's too small to make out properly on the above graph but interstate sides ended up half a game in front of 50% that year too (same in 2020), with no exact 50% wins year season so far. As to what the actaul venues of these home and away games were I can't attest.)
Anyways I know what you're thinking, crazy interesting as that was, what was the actual result of all those games? How did those extra wins/losses translate onto that year's ladder? That brings us to the next graph...
You'll notice I graphed the 'expected' ladder positions, these values are based on the idea that in an 18 team competition that is "completely even" a group of say, 8 teams, selected "at random" would be expected to between them place 1st 8 times every 18 years. If you tell a computer to work backwards for you on those changing numbers, this is what you end up with.
This was one of the more surprising results of the whole thing; positions 1-4 and 6-7 are bang on their expected values. The sum of frequencies of actual position placemet is only ~0.04 in front on the sum of expected frequencies. If we have a look at the differences between expected and actual position placement for non finals positions (6: -1.274, 7: 0.926, 8: -1.074 (6-8 only takes into account the years they weren't finals positions), 9: -6.627, 10: 2.373, 11: 1.373, 12: 0.373, 13: 2.789, 14: -0.211, 15: 0.647, 16: -2.287, 17: 1.588, 18: 1.0) it turns out the impact of that extra home and away loss was felt by the poorer interstate sides in a year, and caused interstate sides as a group to miss out on finishing 9th rather than miss out on qualifying for finals.
Interstate sides are leaning a tad more 5th than 8th, but I don't read much into that. (Since the introduction of the current finals system, those remain the only two positions to never make the grand final, and in their annual showdown 5th is only 1 win in front of 8th! The record of 11-10 defies logic (imo 5th seems to be usually not too far off the top 4 and 8th usually just scapes/scraps into finals, so idk how this is the case)).
I could have just looked at the end total differences for finals sides as well instead of graph them, but I wanted to see if they were advanving at the expected rate or if the overwhelming majority of the progress happenned in a one off period. Most of them were advancing at the expected rate, especially 5th and 6th, but you can clearly see that between 2001-2007 (~25% of the total time (technically more like 18-19% but I added a bit, cos you couldn't expect many placements in early years; the per year number of interstate sides expected to be qualifying for finals didn't pass 2 till '94 (currently at 3.6 btw))) 50% of the total progress of 2nd, ~46% of 3rd, ~38% of 1st and 36% of 8th occurred.
Anyways, so interstate sides are qualifying for finals alright, how are they doing? How has a game being a final affected the win rate historically?
...positively it turns out, as funnily enough this is virtually bang on the expected value (50%) too.
But this doesn't show/capture which finals are being won. How much of that 50% are the big important finals and how much are the cheapy dud ones?
As you can see from the slight curves to the lines, over the years as more sides are added to the comp expected interstate side premierships have been accelerating as a group, while slowing down individually. The expected group premierships is currently going up ~0.45 premierships a year (which would be ~0.9 prelim wins (grand final appearances) a year) (and individual premierships going up ~0.055 a year), so at the end of this season we'd need an all interstate grand final to make the cosmic ledger balanced again for premierships and grand final appearances. (West coast expected premierships same as Brisbane's by the way, I just can't figure out how to alternate lines so they get drawn on top of each other, apologies (also apologies for grouping the bears and post merger lions together, but I felt they fulfilled the same function in this context)).
Anyways interstate sides have never been 2 premierships below nor 4 premierships above the expected line, and have spent 21 total seasons below the line and 18 above.
I feel I should point out here that I actually am heavily against the unfair scheduling of grand finals and prelims at the 'g, it is crazy rediculous. Worth pointing out here that this graph only has 12 data points to compare and work with, compared to the one and a half thousand in the first graph, which is my guess for why interstate sides aren't further behind than they are. I'm sure the FIXturing will definitely show up in the stats before (if) the MCG grand final contract ever expires...
Also worth noting that although the group is on track so far, individually it's more whacky. A year is a crazy long time, and with 18 teams each side is expected to lose a grand final, win 1 premiership than go into an immediate 16 year grand final draught before repeating again. So while Adelaide have a 22 year premiership drought so far (second longest of an interstate side after Freo), they're still about 3 years ahead of their expected premiership schedule, which means they've only just passed the halfway mark of their expected premiership drought (assuming the AFL doesn't add any more teams in the next 20 years, which would extend it). Meanwhile of all things Gold Coast is considered the second most hard done by interstate team on expected premierships (after Freo). Port then are the least hard done by of the sides in the -ve, depite having a premiership drought 7 years older than Gold Coast has been a thing.
All data used in graphs from (and much thanks to) afltables.com, this whole thing ended up being a much more fun and a more fantastic waste of time than I had hoped, learned a bunch, and I hope you found this (or at least some of it) as interesting as I did. Also apologies for the essay this turned out to be .
Last edited: