Aussie Rules World Cup Discussion Thread (OOC)

[Sarzonia]

Well you clearly do have at least a bit of an idea! At least more than most people, and probably myself included

Well, I'm drawing from my experience of never having seen women's lacrosse before and then going to cover a match during my newspaper days. And RPing field hockey and winning twice there.

I couldn't blame the editor about the women's lacrosse thing because I WAS the editor!

[Sarzonia]

Great job hosting by Sylestone and Baggieland! Fun competition to RP!

Also, I got further inspiration to RP, but posted it in my own newsfeed rather than clog the RP thread.

[Sylestone]

Congratulations, Sarzonia!
You certainly deserved the win.
It was nice to see that two nations who had never made it to the semifinals before make it, which is good for the sport as a whole.
The updated ranks are in my signature. If anyone wants to suggest any changes, then I’m all ears.
After next cycle, my presidential term will end, and if anyone wants to put their hand up to run for president or vice-president, make sure you keep it in mind. We can start the process of nominations when signups for the next ARWC open (I’m tentatively thinking August-September)

[Sarzonia]

Thanks Sylestone!

I am curious about the weight a championship final has when compared to the other matches. It seems odd that I'm ranked second with two championship wins behind the team I beat in the most recent final, even though the margin is razor-thin.

[Baggieland]

Hmmm... I'm kinda curious as to why I'm the ONLY nation to have dropped points. I was expecting to lose a bit of ground to Sylestone and Sarzonia, and I was expecting ranks 4 to 9 to have caught up a bit with me. But I was also expecting to have increased my kbp a tad.

[Sarzonia]

Hmmm... I'm kinda curious as to why I'm the ONLY nation to have dropped points. I was expecting to lose a bit of ground to Sylestone and Sarzonia, and I was expecting ranks 4 to 9 to have caught up a bit with me. But I was also expecting to have increased my kbp a tad.

I think you probably slipped after losing in the quarterfinals whereas you made it to the finals in each of the previous editions.

[Sylestone]

Thanks Sylestone!

I am curious about the weight a championship final has when compared to the other matches. It seems odd that I m'm ranked second with two championship wins behind the team I beat in the most recent final, even though the margin is razor-thin.

This is one my main concerns, too.
This is because both of us both dropped a game (yours against HUElavia, mine in the final against you). However, if people want, I am willing to change this.
Whether this means increasing the weighting on the final, or even going to the extent of ensuring the champions get full ranking points, no matter how many games they lose. If we're going to change it, my preference is probably the second option.
Any thoughts?

[Ko-oren]

The second option sounds good to me too. It's what happens in the World Bowl, I believe. What I'm more concerned about is the sheer distance between the top 3 and the rest, and the distance between the top 6 and the rest. I can't tell whether a well RPing team from 4th-6th can even contend for a final, and I don't know if the teams ranked 7th+ can get into a quarterfinal/semifinal at this point. The redeeming factor is obviously that the top three are RPing very consistently and very well, so the success is 100% deserved, but I'd hate to see a tournament that's so young yet so 'stratified'.

[Baggieland]

I think you probably slipped after losing in the quarterfinals whereas you made it to the finals in each of the previous editions.

Ko-oren also slipped in the quarters after making the semis in the previous editions, Ko-oren gained two points, shouldn't that have gone down a tad according to this formula?

What I'm more concerned about is the sheer distance between the top 3 and the rest, and the distance between the top 6 and the rest. I can't tell whether a well RPing team from 4th-6th can even contend for a final, and I don't know if the teams ranked 7th+ can get into a quarterfinal/semifinal at this point

This concerns me too, the big three could easily coast-rank to the semis every time. (As long as they don't face each other in the quarters!).

[Sylestone]

The second option sounds good to me too. It's what happens in the World Bowl, I believe. What I'm more concerned about is the sheer distance between the top 3 and the rest, and the distance between the top 6 and the rest. I can't tell whether a well RPing team from 4th-6th can even contend for a final, and I don't know if the teams ranked 7th+ can get into a quarterfinal/semifinal at this point. The redeeming factor is obviously that the top three are RPing very consistently and very well, so the success is 100% deserved, but I'd hate to see a tournament that's so young yet so 'stratified'.

Yeah, I'm also concerned about this too. However, no way comes immediately to mind about how to fix this apart from having huge RP bonuses. It's not helped by the scorinator itself seeming to favour rank over randomness (although there were a few shock results this cycle, notably HUElavia/Sarzonia and Hebitaka/HUElavia).
I'm also worried about the lower-ranked teams not being able to improve. However, Hebitaka did make it to the semifinals this time, so maybe it's not that big of an issue, although I'm thinking that that's not the case.
As for giving the champion full ranking points, I'm all ears to this suggestion. If enough of the Aussie Rules community wants a fix, it's easily changed.

[Baggieland]

Dear Aussie Rules community;

As we know, there were some grumblings after the ranks and kbp were posted after ARWC III.

These were:
1. How is Sarzonia ranked 2nd when he's won two out of three tournaments?
2. Why did Baggieland drop kbp points when Ko-oren gained points? (Note: I am in NO WAY trying to get points deducted from Ko-oren, just trying to find a way that seems fair and equal to all players!).
3. Sarzonia and Sylestone (and to a lesser extent, Baggieland) are running away with the kbp points, meaning that future editions will be a foregone conclusion.
4. How does Juvencus (only posted a roster) be higher up than Megistos (two-time entrant, RPed a bit, quarter-finalist).

Myself and Sylestone had a chat regarding the above points, it started with me mostly moaning about point two above , but evolved to encompass all the points above. The entire conversation is spoilered below.

After reading the conversation and looking at the proposed changes, myself and Sylestone would really appreciate everyone's thoughts and opinions. All feedback is welcome.

Baggieland
https://docs.google.com/spreadsheets/d/ ... sp=sharing
I've created the above document as a proposal for an alternative way of calculating the rankings. The advantage of this way is that you don't need to be a mathematical genius to understand it; it's straightforward.
Of course, I'm not saying that we MUST use this, it's a suggestion only. There are a couple of discrepancies with the current rankings: Sarzonia is just above you (well two times world champion has to count for something), Ko-oren retakes 4th spot from TJUN-ia, finally MU and Megistos are ranked above CAP and Juvencus, which makes sense to me as the former played in editions 1 and 2 whereas the latter only played in edition 3.
I've created this as I still don't get why I'm the ONLY team to have dropped points when I was unbeaten in the group stage and made it into the knock out stage. As I've already said, I was expecting to lose a bit of ground to you and Sarzonia and I was expecting ranks 4 to 9 to catch up with me a bit. However, I was also expecting my kbp to increase a teeny tiny bit due to the group stage results.
Anyway, have a look at the document and let me know how you feel about a revision of how kbp is calculated.
P.S. I've just added up the wins, both you and Sarzonia have exactly 20 wins each from the three editions, so the fact that he's two time champions seems to me that he should definitely have the first rank.

Sylestone
Yeah, it looks good.
I can explain the formula to you if you want.
My rankings are calculated using the same formula that the KPBs are. You get x4 points for the most recent edition, x2 for the second most recent and x1 for the third most recent.
The reason why you dropped is that you've done so well in previous ARWCs that when you got knocked out in this edition, their inevitable drop of points actually brought you down if that makes sense.
The reason why I'm ahead of Sarzonia is because we both actually got the same number of points from this edition (both lost one game). However, I am willing to change this to make the champion have maximum ranking points.
My only issue with your rankings is that there doesn't seem to be any weighting on the most recent edition. In fact, because ARWC I was longer, it had a lot more weighting than the other two editions.
Other than that, though, it seems pretty good at first glance. Maybe we can put it up to a vote, or you can also put your rankings on the discussion thread.
Hope this helps!

Baggieland
"My rankings are calculated using the same formula that the KPBs are. You get x4 points for the most recent edition, x2 for the second most recent and x1 for the third most recent".
Why is edition 1 any less important than edition 3? (and so on).
"The reason why you dropped is that you've done so well in previous ARWCs that when you got knocked out in this edition, their inevitable drop of points actually brought you down if that makes sense".
Not really. So unless I make the final of all future editions, I'm going to drop like a stone?????
"Maybe we can put it up to a vote, or you can also put your rankings on the discussion thread".
I'm just trying to find a better formula that, in the future, won't have complaints like those of mine and Sarzonia's. The current formula is okay, but obviously it needs tweaking somewhat.
"The reason why you dropped is that you've done so well in previous ARWCs that when you got knocked out in this edition, their inevitable drop of points actually brought you down if that makes sense".
Why didn't Ko-oren drop? He did worse than previously, yet his score increased by two whole points???
"My rankings are calculated using the same formula that the KPBs are. You get x4 points for the most recent edition, x2 for the second most recent and x1 for the third most recent".
Megistos played in two editions and reached the knock outs. Juvencus only posted a roster. How are they better than Megistos?
Sorry to keep hassling you!

Sylestone
I also agree that something needs to be changed, and you have come up with some good suggestions. I actually appreciate you trying to hassle me if you get what I mean.
1. "Why is edition 1 any less important than edition 3? (and so on)."
There are a few reasons why the most recent edition is more important than the rest. One of these is because it helps give newer nations a boost, knowing that if they have done well it will count and it also reduces the advantage that nations get when they have participated for a longer time. Another thing is that when the third most recent tournament drops off the rankings, nations who did really well in that one will have an inevitable drop even if they did well in the previous one, while others who didn't do so well will have quite a large increase, even if they don't do particularly well. An example of this is in lacrosse, where WLC 36's runner up dropped from fifth to seventh, while the fourth-placed side jumped from eighth to third despite doing nowhere near as well. So, essentially, it is to ensure that recent success is to be awarded. A very good question though - it took me quite a while to put all of that together.
2. "Not really. So unless I make the final of all future editions, I'm going to drop like a stone?????"
It will be a gradual drop, but possibly. However, if you don't do so well, then you will drop on any sort of ranking formula. Sometimes you just get bad luck, and that's the way it is. It will happen to all of us at some point, you've just got incredibly unlucky. The only reason why Ko-oren didn't drop but you did is because Ko-oren, while they've done well, they haven't done as well as you have. They were ranked fourth after the first edition and then only made the quarterfinals in ARWC II. Because they did that again this cycle, they gained ranking points as they still did pretty well. You, on the other hand, have nearly got full ranking points in both of the first two editions, so when you were eliminated early on by Sarzonia, you dropped. I hope that makes more sense.
3. "Megistos played in two editions and reached the knock outs. Juvencus only posted a roster. How are they better than Megistos?"
This is a very good question, too. This is due to the weightings. But if you look at the points they have got, they have only just made the knockouts on each occasion (often because other results went their way) and been knocked out in the quarterfinals. However, it is a very good question and maybe something that we can look into.
A possible solution could be only having the most recent edition being weighed x3, then x2 and x1?
I really don't know. There doesn't seem to be any "perfect" solution.
But I really appreciate that as a community (and a lot of this is by you, which is really appreciated), we're trying to make this sport a lot more even and not as stratified.

Baggieland
I spent a good part of yesterday researching RL sports ranking formulas: football, rugby, cricket, tennis, golf. What I found was that all of them were extremely complicated and they all had controversies and criticisms. So, if professional people who are paid to create ranking formulas can't come up with a perfect system - then we're in good company!
Onto our conversation from yesterday. I'm now okay with the fact that Baggieland dropped points. In RL, Uruguay won the football world cup in 1930 and 1950, but have done nothing since then. Uruguay can't expect to be in pot one for eternity based solely on what they did 70 years ago. And neither can Baggieland for that matter!
However, the current system definitely has its flaws. Sylestone is above Sarzonia despite the latter being two times champion - that can't be right! If Baggieland dropped points for doing worse than previously, then how does Ko-oren gain a whopping two whole points when they didn't improve on previous editions? A slight increase, maybe, but not that large amount. Finally, there's that Megistos - Juvencus discrepancy.
I've tweaked that google sheet I sent you yesterday to give a little more emphasis on recent editions and a little less emphasis on earlier editions. Remember, the ranking formula is based on the last three editions only. So when ARWC IV is over, the pre-tournament ranks for edition V will be based on ARWC II, III and IV. Edition I will then no longer factor into any rankings (and so on).
The rankings under my proposed system are the same as we have now, except that Sarzonia is now No. 1, and rightly so. Ko-oren retakes 4th spot from TJUN-ia (both these teams have made two quarters and one semi each, however, Ko-oren has won more group matches than TJUN-ia, so this feels right to me). Finally, Megistos and Juvencus are even.
I'm sure you'll notice that under my proposal the rankings will always be between 0.00 and roughly about 15.00. I think this is better as it means that no teams can get so far out of sight that no one else will bother with this tournament. Take the semi final between Sarzonia and Hebitaka. Sarzonia's pre-tournament KBP plus the extra points he gained RPing during the tournament meant that Hebitaka never stood a chance. Under this system, as long as Hebitaka RPs regularly, that semi would have been a lot closer.
I would like to share this proposal with the Aussie rules community, however, I'm a bit worried that some will just simply say "That's not the way KBP points are done". So, I'd really appreciate your thoughts first.
The link to the sheet is the same as yesterday. Just in case:
https://docs.google.com/spreadsheets/d/ ... edit#gid=0

Sylestone
I think that this is an almost perfect compromise. My only issue is the same weighting on the knockout matches, but that's such a minor issue that we would need more people from the community to back us up on that one.
As for the people saying that "That's not the way ranks are done", I'm on your side. There shouldn't be any limitation to the way rankings are calculated, just as long as they make sense. And as this makes more sense than my current rankings, I'm all for it.
I say go for it.
Do you think it would also be handy to maybe release this conversation onto the discussion thread (maybe under a spoiler)?

[Ko-oren]

I only just read through it, a lot of good ideas there.

First: 'KPB' refers strictly to the association football rankings (and its system) afaik. For ARWC, I'd just say 'rankings' or 'coefficients' or something like it.

Another ranking system we could steal is American football's (here) which give out full points for the winner, and gives out points to the other teams in the following formula: (W/TGP)*MOD where W is wins, TGP is total games played (by that team), and MOD is a modifier (15 in the World Bowl, for all of the past 3 tournaments). Results from the last edition count 100%, results from the edition before that count for 50%, and from the edition before that count for 25%. I think Sarzonia has some experience with that system as well, and could weigh in or correct what I said.

I support halving the points of last ARWC and dividing them by 4 for the one before that, for the reasons Sylestone gave.

There's not really a single way in which rankings are best calculated: even the KPB (Kaze-Progressa-Bedistan rankings) aren't the be all and end all, necessarily. 'But that's how we've always done it' is an abysmal reason for doing something in my book, so I'm with you. For other sports we're free to come up with what we need. I'm fine with the system Baggieland proposed, but I'd like to see KO wins in previous editions also halved/quartered (as Syle noted at the end). The system as it's in the spreadsheet now rewards teams disproportionally for winning knockout games, making it more and more unlikely for new teams to join the group of knockout winners - because once you've won one, you've got a full point for the next three three tournaments, which I feel is a bit generous towards the top 4/8. You're already rewarded for winning more group stage games, I don't see the need to doubly reward these teams for also making the knockouts (and winning one or two games).

[Baggieland]

Thanks a lot for your input Ko-oren.

So, the best way to compare Ko-oren's suggestion with what we had already is to put it into that google document.

https://docs.google.com/spreadsheets/d/ ... edit#gid=0

(same link as before, Ko-oren's suggestion has been added to the bottom, original suggestion is at the top).

Most notable observations are that Sarzonia and Sylestone are now dead equal. TJUN-ia retakes 4th spot from Ko-oren.

[Sarzonia]

I definitely support a proposal to revise the rankings, if for no other reason than the prospect of a two-time champion being ranked behind the team it defeated in the most recent final seems rather ridiculous.

Even having the two teams dead even in the rankings doesn't seem right.

I'm not as well versed in the mechanics of rankings and the formulae used to create them, but I think I would be more likely to support Baggieland's compromise given that it elevates me in total points given what I said above.

I also share the concern that the gap between the top three and the other teams makes it more difficult for other teams to break into the top tier. Perhaps slight increases in weight for only the 3PPO and the final might help?

[Ko-oren]

The reason you and Syle even out in the second system is because Syle won more group stage games in the lastest edition, as well as winning the second most recent tournament, which is still a decent chunk of points. I feel the second system is balanced a little better than the first one: it reflects Syle's dominance in the group, though I'll admit it's uncanny how you don't take first place outright. The fact you won the first edition is (almost completely) moot, because those points are divided by four: three group stage wins in the most recent tournament give 2.25 points, more than what you get for winning a tournament two editions ago. I feel that's fair.

One last thing I thought of: points are given out per win. What's stopping me from hosting an edition with a full round robin group stage, massively increasing the amount of ranking points in the system? I'd like to again propose something like Wins/Matches*Modifier which rewards teams fairly across different tournament formats.

[Baggieland]

I definitely support a proposal to revise the rankings, if for no other reason than the prospect of a two-time champion being ranked behind the team it defeated in the most recent final seems rather ridiculous.

Even having the two teams dead even in the rankings doesn't seem right.

I'm not as well versed in the mechanics of rankings and the formulae used to create them, but I think I would be more likely to support Baggieland's compromise given that it elevates me in total points given what I said above.

I also share the concern that the gap between the top three and the other teams makes it more difficult for other teams to break into the top tier. Perhaps slight increases in weight for only the 3PPO and the final might help?

Thanks Sarzonia, I just spent a while fiddling around with that table, mostly by giving weight to the latest edition - then halved - then quartered for editions I and II. However, every change I made had Sylestone coming out above you! So, it seems the first one I did is closest so far. Regards extra weight for 3ppo, I figured that in too, but I felt the winner of 3ppo shouldn't get more points than the losing finalist. Second place is better than third place, right!

One last thing I thought of: points are given out per win. What's stopping me from hosting an edition with a full round robin group stage, massively increasing the amount of ranking points in the system? I'd like to again propose something like Wins/Matches*Modifier which rewards teams fairly across different tournament formats.

Hey, Ko-oren, I'm not great at maths either, I can do addition, subtraction, multiplication and division, but formulae gets me muddled up. Can you actually show us how the "wins/matches/modifier" looks like? Once I see an example rather than a formula, I'll probably understand it better. Don't forget to show your working out!

[Ko-oren]

One last thing I thought of: points are given out per win. What's stopping me from hosting an edition with a full round robin group stage, massively increasing the amount of ranking points in the system? I'd like to again propose something like Wins/Matches*Modifier which rewards teams fairly across different tournament formats.

Hey, Ko-oren, I'm not great at maths either, I can do addition, subtraction, multiplication and division, but formulae gets me muddled up. Can you actually show us how the "wins/matches/modifier" looks like? Once I see an example rather than a formula, I'll probably understand it better. Don't forget to show your working out!

I hope the permissions are right and everyone can read this: 2 proposals.

Possibility 1: (Wins/Games*10) 10 is arbitrarily chosen and only serves to make the final product read like "15.515" rather than "0.15", as let's be honest, larger numbers look better. In this case, as ARWC I consisted of 9 games while II and III took place across just 7, the first ARWC isn't disproportionally rewarded (because in ARWC I, you could've picked up 9 wins, while in the others you could only win 7. What's stopping me from making a 20-game ARWC and ruin the rankings for the next 3 editions?). Note the columns D, G, and J, calculating the rank per ARWC. Column K is the final product, dividing ARWC I rank by 4, ARWC II by 2, and awards full points for the most recent edition. As you can see, Sylestone ranks first given their overall better performance: they were better (flawless, even) in ARWC II while they were even with Sarzonia in ARWC III.

And now you see a flaw in Possibility 1: it doesn't reward tournament winners. Cue Possibility 2.

In Possibility 2, it's still (Wins/Games*10), with all columns working exactly like Possibility 1. The one exception is that tournament winners are given 1 win from 1 game, representing a flawless tournament, and awarding them full points for that ARWC regardless of any group stage loss. In this system, Sarzonia ranks first by winning two editions (and thus getting full rank from these) and only losing once in ARWC II. Sylestone ranks second.

I'd like to propose Possibility 2.

Pros:
+ Makes long tournaments and short tournaments yield the same amount of max points, instead of rewarding longer tournaments.
+ Degrades rank over time and awards more points for more recently completed ARWCs.
+ Rewards tournament winners.

Cons:
- The 2nd and 3rd placed teams often get the same amount of points, as they both finish their campaigns with 1 win and 1 loss. This makes them look like they had similar finishes though you could argue that finishing 2nd is better than finishing 3rd.
- Slightly more complicated.

(You can solve the first con by overwriting both the winner's (-> 1/1) and the runner-up's (-> [games -1]/[games]) points. In this way, Sarzonia would get 1/1 from ARWC III, and Syle would get 6/7 regardless of how many group stage games they lost.)

[Sylestone]

Apologies for not being in the conversation earlier on, I just felt as though I couldn't really contribute!
With Ko-oren's second proposal, I also like it - I think it provides a middle ground between what has been suggested between Baggieland and me, along with the feelings of the rest of the community. However, I might also like to perhaps suggest a few minor changes in relation to the first con that Ko-oren has pointed out.
I have noticed that there is a possible solution, with the runners-up always earning only one fewer wins than what the winner ends up having (full points), which might solve the issue with the second and third places having equal rank, which should be open to consideration.
However, there could also be another solution. We may not even need to have any sort of bonus for finishing third, partly because ARWC I didn't even have a third-place playoff. Or, instead, we could only give the winner of the 3PPO half points for the win, in comparison to other wins across the tournament.

Sorry if this comes out as gibberish, I've only just woken up.

[Baggieland]

Yeah, Ko-oren's 2nd proposal is close to what everyone wants, me thinks!

partly because ARWC I didn't even have a third-place playoff

That was a one-off and shouldn't happen again. Once ARWC IV is over, it will no longer be a consideration.

[Baggieland]

I've updated Ko-oren's 2nd proposal just a teeny tiny bit, only in order to make the maths / working out obvious.

(same link as before).

https://docs.google.com/spreadsheets/d/1n0P0zlk9_Q2w7lCUBnedjExeJyj4xMOJju03isT6CjA/edit#gid=0

Could we have a vote on this?

[Ko-oren]

I'm ok with this system. I think the protection for the tournament winner should be in place. The way I see it we have three choices:
- The system as in the post above (1st place protected)
- The system as in the post above, but with added protection for the 2nd place finishers (making sure they get the ranking points equivalent to a team losing one game all tournament. Mentioned by me and Syle as a potential downside of the system as above)
- Decline the system as in the post above

(In that case, I'd vote for option 1)

[Sylestone]

I'd probably vote option two. I like the idea of having runners-up protection.
If we get someone else to weigh in, I'll organise a formal vote with all the nations who had rostered in the last edition (apart from puppets, of course).

[Baggieland]

Okay, I think we all need to see what guaranteed points for a second place finish looks like before we vote on it. How many points? If first place gets 10.00, how about 9.00 for second place?

[Baggieland]

I've gone ahead and added another table with 9.00 for second place.

Obviously, at the moment, this only affects Sylestone and Baggieland. Resulting in roughly 0.50 Total increase for both these nations.

https://docs.google.com/spreadsheets/d/1n0P0zlk9_Q2w7lCUBnedjExeJyj4xMOJju03isT6CjA/edit#gid=0

[Sylestone]

Okay, I think we all need to see what guaranteed points for a second place finish looks like before we vote on it. How many points? If first place gets 10.00, how about 9.00 for second place?

Could be. It will be full points for all matches, save the final. However, this is only if there at ten matches in the competition (for the winners and runners up, as well as third and fourth place). If there are say, seven matches, then the runners up will get 6/7 * 10 points (8.86...)