Under a 20 team league, construct a 10 team final series

[9874123]

I posted this earlier in the thread, but when you take these benefits into account, the mathematical probabilities of different systems look like this:

	Finals system
Ladder position	Current system with 7v10, 8v9 wildcard	2x McIntyre final 5	Knockout final 10
1	21.5%	22.8%	18.5%
2	21.5%	22.8%	18.5%
3	19.2%	12.5%	14.6%
4	19.2%	12.5%	13.5%
5	6.0%	10.7%	10.2%
6	6.0%	10.7%	9.6%
7	1.8%	2.3%	4.3%
8	1.8%	2.3%	4.3%
9	1.4%	1.7%	3.2%
10	1.4%	1.7%	3.2%

So even in the knockout system with no double chance, the top team has roughly 6x the probability of winning compared to 10th, assuming all teams are evenly matched.

In the other systems it's 13-15x. But the advantages are not as nicely distributed.

Appreciate the work put into the post but I think we will just have to agree to disagree. I don't think 1st being 6x more likely to win the premiership than 10th is a strong enough advantage. It would mean a team from 7th/8th/9th/10th cumulatively wins the premiership almost as often as the minor premier. For what its worth, I ran a simulation using the same assumptions (Home team has a 65% chance to win off a bye, home team has a 57% chance to win if not off a bye, GF is 50/50) and this is what I got.

	Finals system
Ladder Position	ibd77
1st	24.6%
2nd	24.6%
3rd	13.1%
4th	13.1%
5th	9%
6th	9%
7th	1.8%
8th	1.8%
9th	1.4%
10th	1.4%

Its going to be up to personal taste as to which set of numbers one prefers. I prefer these numbers. One could argue that it doesn't give 7th onwards enough of a chance, but to be honest, I don't really want to see a team that couldn't put together a consistent season winning the premiership too often.

 

[9874123]

Given 4th has only made the GF 3 times, this also suggests they are typically the easier opponent in the PF, yet the top side never plays them in the PF, instead typically playing 2 or 3.

I would argue that 1st making the GF 17/24 times and 4th making it 3/24 times suggests, as I said earlier, that it is much more of an advantage to play 4th in the Qualifying Final rather than the Prelim, since winning the QF is such a huge advantage, giving you a week off and home ground advantage in the Prelim.

Not sure if you could say definitively that having beaten a stronger opponent the week before leading into a GF is a disadvantage. Maybe it is, maybe its an advantage, since it prepares you better. Iron sharpens iron and all that. But given that the GF is a whole other mess of home ground advantage for lower ranked teams, vic vs interstate clubs etc that its hard to apply the same logic to reach a solid conclusion in my opinion.

 

[NeutralFan]

Okay but what score did OP get for their assignment?

 

[Chiz]

Appreciate the work put into the post but I think we will just have to agree to disagree. I don't think 1st being 6x more likely to win the premiership than 10th is a strong enough advantage. It would mean a team from 7th/8th/9th/10th cumulatively wins the premiership almost as often as the minor premier.

Keep in mind these numbers assume all teams are evenly matched. A team from 7-10 is rarely going to win if they are genuinely worse than the top sides (which will be the case most years). On the off chance that a 7-10 team is carrying the most form into the finals, they should still have to work hard to win, but it should at least be realistic.

I also note you originally referred to the system needing to provide adequate incentive for teams to finish higher. The knockout system does this in my opinion, not only between 1st and 10th, but between each set of two ladders positions. The current system groups teams into four with insufficient incentive to finish higher within the four (particularly given home ground advantage is not always a thing).

For what its worth, I ran a simulation using the same assumptions (Home team has a 65% chance to win off a bye, home team has a 57% chance to win if not off a bye, GF is 50/50) and this is what I got.

	Finals system
Ladder Position	ibd77
1st	24.6%
2nd	24.6%
3rd	13.1%
4th	13.1%
5th	9%
6th	9%
7th	1.8%
8th	1.8%
9th	1.4%
10th	1.4%

Its going to be up to personal taste as to which set of numbers one prefers. I prefer these numbers. One could argue that it doesn't give 7th onwards enough of a chance, but to be honest, I don't really want to see a team that couldn't put together a consistent season winning the premiership too often.

Appreciate engaging with someone who understands the numbers! Personally I prefer the knockout system numbers, but we can agree to disagree. I think this system works better than the current one in terms of distributing the chance across the ladder position.

The huge drop between 6th and 7th is an issue though - keep in mind the Bulldogs and Hawthorn this season are separated by percentage only, and this system would give a 5x advantage to the higher-placed Bulldogs.

 

[ibd77]

This is my favourite system here so I drew it up to make it easier to visualise. Home team is first, and assumes the higher ranked team wins each week. I like it better than simply adding a 7 v 10, 8 v 9 wildcard weekend because it further benefits teams in the top 2. In the current system, there is a huge difference between finishing 4th and 5th. In this system, the advantages are evened out a little between every 2 position drop.

In this system:
- finishing 2nd is much much better than 3rd
- finishing 4th compared to 5th is basically the difference between finishing 2nd and 3rd in the current system
- finishing 6th rather than 7th is comparable to 4th and 5th in the current system


This system may be a little confusing for the casual fan though.

View attachment 2092675

I used pretty much the same system the SANFL uses for their top 5, just double the amount of games. The top 2 places needs to be rewarded at the start of the finals series. The highest ranked team through out the finals series should get the benefit of extra days break between PF & GF.

 

[9874123]

Keep in mind these numbers assume all teams are evenly matched. A team from 7-10 is rarely going to win if they are genuinely worse than the top sides (which will be the case most years). On the off chance that a 7-10 team is carrying the most form into the finals, they should still have to work hard to win, but it should at least be realistic.

I also note you originally referred to the system needing to provide adequate incentive for teams to finish higher. The knockout system does this in my opinion, not only between 1st and 10th, but between each set of two ladders positions. The current system groups teams into four with insufficient incentive to finish higher within the four (particularly given home ground advantage is not always a thing).

Appreciate engaging with someone who understands the numbers! Personally I prefer the knockout system numbers, but we can agree to disagree. I think this system works better than the current one in terms of distributing the chance across the ladder position.

The huge drop between 6th and 7th is an issue though - keep in mind the Bulldogs and Hawthorn this season are separated by percentage only, and this system would give a 5x advantage to the higher-placed Bulldogs.

I think our fundamental disagreement is that you seem to want the chance of winning the flag to be a smoother distribution between each position on the ladder, where I am happy with a stepped outcome. I don't really see a huge drop between 6th and 7th as an issue. It makes it something worth fighting for, and ensures the H&A season is sufficiently meaningful.

After all, we see teams finish 9th by percentage all the time, and that means they have infinitely less chance of winning the flag than 8th.

I think I've said all I can say here. I appreciate the back and forth, I think it was a good discussion

 

[Chiz]

I think our fundamental disagreement is that you seem to want the chance of winning the flag to be a smoother distribution between each position on the ladder, where I am happy with a stepped outcome. I don't really see a huge drop between 6th and 7th as an issue. It makes it something worth fighting for, and ensures the H&A season is sufficiently meaningful.

I certainly like the smoother distribution, but as discussed I also think there are other fundamental issues with having double chances.

After all, we see teams finish 9th by percentage all the time, and that means they have infinitely less chance of winning the flag than 8th.

True, but you have to draw the line somewhere - in relative terms the difference is infinite, but in absolute terms it's not high given 8th only has a small chance.

I think I've said all I can say here. I appreciate the back and forth, I think it was a good discussion

Likewise, appreciate the discussion.

 

[steve-o]

Okay but what score did OP get for their assignment?

I regret dredging this thread up from the depths.

 

[LordLucifer]

Found it - sorry I'm almost a decade late.

Top 3 get the double chance, and there's a semifinal in Week 2 that only gets played if there's a big upset in Week 1.

Prelims & Eliminators:
A - 2nd vs. 3rd (double chance)
B - 1st vs. 10th (double chance for 1st only)
C - 4th vs. 9th
D - 5th vs. 8th
E - 6th vs. 7th

Errr ............. what ????

 

[LordLucifer]

View attachment 2092675

The only thing I like about that is your graphics skills.

 

[LordLucifer]

Even if the number of teams increases to 20, keep the Final 8 structure.

When the number of clubs grows to 22 (and it will), only then should there be a move to a Final 10 structure.

 

[steve-o]

Errr ............. what ????

If the minor premiers beat the 10th seed in week 1, they get a bye in week 2. The 10th seed doesn't get this luxury if they win that game.

The result of this game determines whether there is a third week 2 semifinal or not - there will either be 6 teams left (2 byes) or 7 teams left (1 bye).

This oddity makes the whole thing work right: same number of weeks, clear incentives for finishing higher up the ladder, and rest advantages/disadvantages in week 3.

Chew on it for a bit before you categorically reject - it's not complicated but it is unlike anything else being discussed. It's the reason I bumped the damn thread - I thought I found the way to do 10 teams that would please nearly everyone, and wanted to share.

 

[LordLucifer]

If the minor premiers beat the 10th seed in week 1, they get a bye in week 2. The 10th seed doesn't get this luxury if they win that game.

The result of this game determines whether there is a third week 2 semifinal or not - there will either be 6 teams left (2 byes) or 7 teams left (1 bye).

This oddity makes the whole thing work right: same number of weeks, clear incentives for finishing higher up the ladder, and rest advantages/disadvantages in week 3.

Chew on it for a bit before you categorically reject - it's not complicated but it is unlike anything else being discussed. It's the reason I bumped the damn thread - I thought I found the way to do 10 teams that would please nearly everyone, and wanted to share.

Too many if's, but's & maybes.

Keep it simple.

 

[steve-o]

One if, no buts, no maybes, and yet you still found it worthwhile to make that comment. I won't try convincing you further, it is a waste of effort.

 

[WesternSydneyHawk]

Here's a possible way to structure it:

First round of the finals:

The teams that finished 1, 2, 3, 4, 5, and 6 all have a week off

7 v 8: Loser eliminated
9 v 10: Loser eliminated

Second round of the finals.

1 v 2: Winner goes directly into the grand final, loser gets a second chance

3 v 4: Winner goes directly into the prelim final, loser gets a second chance

5 v 6: Loser eliminated

(Winner 7 v 8) v (Winner 9 v 10): Loser eliminated

Third round of the finals:

Second chance match: (Loser 1v2) v (Loser 3 v 4)

(Winner 5 v 6) v (Winner 7, 8, 9, or 10)

Fourth round of the finals:

(Winner second chance match) v (Winner 5, 6, 7, 8, 9, or 10)

Prelim final:

(Winner 4th round) v (Winner 3 v 4)

Grand final:

(Winner 1 v 2) v (Winner prelim final

 

[Sydney Kicker]

Sorry I haven't read the whole thread but this may be a different approach...

Rounds 1-19 every team plays each other team once – 9 home matches, 9 away and 1 derby round with a neutral or shared venue (eg WCE v FRE, POR v ADE).

After R19 you split the ladder into a top 10 with 2 groups (A:1,4,5,7,10),(B:2,3,6,8,9) and bottom 10 with 2 groups (C:11,14,15,17,20),(D:12,13,16,18,19).

For Group A and B, each group then plays a round robin over 5 weeks (2h, 2a, 1 bye) against the other 4 teams in their group. Then after a week off the prelims are A1 v B2, and B1 v A2 with the winners to meet in the GF (or you could just go straight to the GF and skip the prelims for a shorter/more cutthroat season). The teams placed 1st and 2nd after the first 19 rounds get the most favorable bye weekend (in the middle of the round robin).

For the bottom 10 each group plays a round robin (2h,2a, 1 bye) against the other 4 teams in their group. Then the winner of group C plays the winner of group D in the free weekend before the prelim finals for the first pick in next years draft (The Harley Reid cup).

As an added equaliser, in the next year you reverse the draw in the first 19 rounds so all the teams you played away in the first year you play at home the second year.

 

[WesternSydneyHawk]

Decided to revise my earlier structure a bit.

First round of the finals:

The teams that finished 1, 2, 3, 4, 5, and 6 all have a week off

7 v 10: Loser eliminated
8 v 9: Loser eliminated

Second round of the finals.

1 v 2: Winner goes directly into the grand final, loser gets a second chance

3 v 4: Winner goes directly into the prelim final, loser gets a second chance

5 v 6: Loser eliminated

(Winner 7 v 10) v (Winner 8 v 9): Home game for highest placed team, loser eliminated

Third round of the finals:

Match 1: (Loser 1 v 2) v (Winner 7, 8, 9, or 10)

Match 2: (Loser 3 v 4) v (Winner 5 v 6)

Fourth round of the finals:

(Winner Round 3 Match 1) v (Winner Round 3 Match 2)

Prelim final:

(Winner 4th round) v (Winner 3 v 4)

Grand final:

(Winner 1 v 2) v (Winner prelim final)

 

[iameviljez]

8 out of 20 is about right IMO.

Don’t forget, it used to be 4 of 12, 5 of 14 and 6 of 16.

 

[Dazzler9]

8 out of 20 is about right IMO.

Don’t forget, it used to be 4 of 12, 5 of 14 and 6 of 16.

It was never 6 of 16. When top 8 first came in there were only 15 teams (1994).

Edit: agree about 8 being enough!

 

[iameviljez]

It was never 6 of 16. When top 8 first came in there were only 15 teams (1994).

Edit: agree about 8 being enough!

Ah whoops - think I might have typo'd. I clearly remember the 8 coming in... because Geelong had finished 7th with a wet sail the year before.

 

[Howard Littlejohn]

Even if the number of teams increases to 20, keep the Final 8 structure.

When the number of clubs grows to 22 (and it will), only then should there be a move to a Final 10 structure.

Eight is enough at 24 teams. Six would still be enough really.
But at 22 teams, promotion-relegation makes more sense. The fixture can be full home and away with 12 top flight teams, there's something to play for at both the top and bottom of the top flight, and top of the second tier.

Yes, the unequal financial distributions would need to be looked at, the second tier would probably need lower salary caps (without being so low that a whole new team needs to be recruited on promotion), and the draft can finally be scrapped. There's nothing unsolvable there while removing the worst of the (before finals) fixture biases.

I know it won't happen; and they would go with play each other once plus a second local rival for extra Collingwood, West Coast and Adelaide money. But from a sporting point of view, it makes more sense.

 

[Piggy Smalls]

If I was making a 10 team finals series I'd have 7 v 10 and 8 v 9 play off while the top 6 have the week off, then revert to the current format with the winner of 7 v 10 playing the team in 6th and the winner of 8 v 9 playing the team in 5th in the EFs. There would be no pre finals bye for the teams 7th to 10th.

 

[ioppolo]

If I was making a 10 team finals series I'd have 7 v 10 and 8 v 9 play off while the top 6 have the week off, then revert to the current format with the winner of 7 v 10 playing the team in 6th and the winner of 8 v 9 playing the team in 5th in the EFs. There would be no pre finals bye for the teams 7th to 10th.

This is exactly how the NBA operates.

Whilst it makes sense if a top 10 were to exist, there simply shouldn't be a top 10 existing in the first place imo.