Register For Our Mailing List

Register to receive our free weekly newsletter including editorials.

Home / 581

The changing face of finals footy and the numbers behind it

Competition in our footy codes drew to a close over the weekend, with Penrith claiming a historic fourth premiership in a row in the NRL grand final. It followed the conclusion of the AFL the week prior, when Brisbane had an emphatic grand final win over Sydney.

Brisbane came from the clouds to win from fifth position on the AFL ladder over the minor premiers. Since the inception of the current AFL final eight system in 2000, this was only the second time that a team has won from outside the top four. The Western Bulldogs won it from seventh in 2016 (with again, the hapless Sydney minor premiers and runners-up).

This got me thinking about the overall performance of top four sides in recent times, where inglorious finals exits seem more prevalent. In particular, I wondered what effect the controversial pre-finals bye introduced in 2016 has had on top four sides’ chances. I had a feeling that teams 1 to 4 were losing continuity in playing, particularly for those teams that won in the first week of the finals, advancing straight through to the third week. Meanwhile, the bye would perhaps favour teams with final ladder positions 5 to 8, enabling them to freshen up before week one of the finals, reset, and build momentum with every final sudden death for them.

First, a quick look at how the AFL final eight system works:

Week 1
Qualifying Final 1: 1st final ladder position vs 4th final ladder position
Qualifying Final 2: 2nd vs 3rd
Elimination Final 1: 5th vs 8th
Elimination Final 2: 6th vs 7th
Week 2
Semi-final 1: loser QF1 vs winner EF1
Semi-final 2: loser QF2 vs winner EF2

Week 3
Preliminary-final 1: winner QF1 vs winner SF2
Preliminary-final 2: winner QF2 vs winner SF1

Week 4
Grand-final: winner PF1 vs winner PF2


Source: AFL.com.au

I looked at two measures to probe my intuition, over the periods 2000 to 2015 when there was no pre-finals bye, and post 2015 when it was introduced:

  1. The frequency with which top four teams won their first final, then exited in their subsequent week 3 preliminary final. So those teams bypass week 2 finals, and in the pre-finals bye period, may have only played one match in nearly four weeks up until preliminary final day.

  2. The frequency with which top four teams exited the finals in so-called ‘straight sets’. That is, a loss in weeks one and two of the finals series.

The results were compelling.

From 2000 to 2015, only four times did top four teams lose the preliminary final having won their week 1 final and bypassing week 2. That is, just 12.50% of the time (4 out of 32 losing preliminary finalists). From 2016 to 2024, seven top four preliminary finalists failed to advance to the grand final after a first week win. That is, a strike rate of 43.75% (7 out of 16 losing preliminary finalists). Note, 2021 has been excluded from the analysis, when there was no pre-finals bye due to a Covid restructured season. 

That’s a significant increase in rate of failure to advance, which supports the theory that top four sides were losing continuity in game time, while the lower ranked teams had momentum on their side.

Straight set exits have also spiked

For the period 2000 to 2015, there were five straight set exits out of a total of 64 top four finals participants, a rate of just 7.81%. For 2016 to 2024, that rate jumped to 21.88%, as seven out of 32 teams lost in rounds one and two of finals. 

Again, this backs the assertion that teams ranked 5th to 8th have closed the gap on the top four. Perhaps because the week off has provided a circuit breaker between a long season and an arduous finals series, where the teams need a win every week to progress. Without the bye, there would be no respite for those teams.

The pre-finals bye was introduced by the AFL in 2016. By that point, there had been several cases of teams that had already locked in a certain finals position resting a number of players to keep them fresh for the upcoming finals. The AFL thought this compromised the integrity of the competition as clubs were not fielding their strongest available teams. In theory, a pre-finals bye meant that teams would not need to sideline players in the last round.

But did the ruling have unintended consequences? The analysis here suggests that it did. The AFL may be comfortable with that as it seems to have brought about a less predictable finals series. Others would argue that it unfairly disadvantages higher ranked teams, who have worked hard over a long season to achieve top four status and the week one double chance.

Note that this year, minor premiers Sydney did not fall into the two categories analysed here, instead winning in weeks 1 and 3 before losing the grand final. However, by the day of the preliminary final, Swans had played only one game in 27 days. Meanwhile, all-conquering Brisbane had played four hard-fought finals in as many weeks, building up significant momentum in the process. Did a lack of match practice - aided by the pre-finals bye - bring about Sydney’s demise?

Footnote

For those interested, assuming each team has a 50% chance of winning (or losing) per final, the probabilities of teams ranked 1 to 4, and 5 to 8 winning the premiership, can be calculated at the outset of the AFL final eight series.

Teams 1 to 4 can become premiers via one of two paths:

  1. Win weeks 1, 3, and 4 (the path most frequent, with 17 premierships since 2000).
  2. Lose week 1, win weeks 2, 3, and 4 (6 premierships).

Therefore, the probability teams 1 to 4 win the premiership = (0.5 x 0.5 x 0.5) + (0.5 x 0.5 x 0.5 x 0.5) = 18.75% each.

Teams 5 to 8 can only become premiers via one path: win weeks 1, 2, 3, and 4 (happened only twice). Therefore, the probability teams 5 to 8 win the premiership = (0.5 x 0.5 x 0.5 x 0.5) = 6.25% each.

Note that these numbers are more about relativities than absolute probabilities because in reality, a probability other than 50% of winning per final would be arrived at by factoring in characteristics like recent form, home ground advantage, injuries, and so on.

The numbers show that teams 1 to 4’s chances of winning the grand final, are possibly several multiples of that of 5 to 8’s chances. Which emphasises the difficulty of winning from outside the top four, and puts Brisbane’s effort into perspective.

 

Tony Dillon is a freelance writer and former actuary. This article is general information and does not consider the circumstances of any investor.

 

11 Comments
Fairness is necessary
October 30, 2024

I think AFL current system is very good, except one thing.
I'm sure that path from Seed 2 is worse than from Seed 1? 
Because the first opponent is stronger (3rd instead 4th), but in knockout stages:
-If better seed always wins, 2 plays prelim against 4 and 1 against 3
-If both lose, Seed 2 plays semi probably against 6 and 1 against 5
This should be analyzed.

Tony Dillon
October 14, 2024

Thanks for doing that Martin. I was going to do it but you saved me the trouble. I note that the NRL numbers are closer to the AFL (no bye) numbers than AFL (bye), which kind of supports my argument. Though the 'no bye’ numbers are not quite as pronounced with the NRL, there may be reasons for that.

One possibility could be that the home ground advantage (for higher ranked teams) may not be as strong in the NRL, so that there is not as much as a natural gap between 1 to 4, and 5 to 8 in the earlier finals rounds. In the NRL, teams only really travel between NSW and Qld, as opposed to all around the country with the AFL. And the NRL has standardised playing areas whereas the AFL doesn’t. For example, the unusually sized SCG would be a big advantage for Sydney insofar as game plans and familiarity are concerned, even maybe types of players recruited. Which becomes a disadvantage for Sydney when it has to play the grand final at the MCG, where the dimensions are much different to the SCG and Sydney hardly plays there. I believe that's another factor behind Sydney's poor recent grand finals record.

And interesting you mention the old “final five” format. I always thought that was the best finals arrangement, as it gave the minor premiers a genuine statistical advantage over the other finalists. Similar to my footnote, relativities in chances of winning the flag can be calculated:

Team 1 premiership paths: WW, or LWW = (0.5 x 0.5) + (0.5 x 0.5 x 0.5) = 37.5%

Teams 2 & 3 premiership paths: WWW, or LWWW, or WLWW = (0.5 x 0.5 x 0.5) + 2 x (0.5 x 0.5 x 0.5 x 0.5) = 25%

Teams 4 & 5 premiership path: WWWW = (0.5 x 0.5 x 0.5 x 0.5) = 6.25%

Martin Mulcare
October 15, 2024

Thanks Tony,
Your thoughts on the relative value of "home advantage" are valid, I think, especially the different shaped grounds. Which reminds me that there was little home ground advantage in much of the "final five" era when most finals were played at the SCG. The "2nd chance" was the only practical means of recognising higher ladder placings.
(And we can save analysis of the McIntyre System for another audience).

Tony Dillon
October 17, 2024

Definitely probably for another audience Martin, but I will say another reason I like the old Final Five format is because all finals are not necessarily sudden death until week 3. Teams 1 to 3 can lose in week 2 and not be eliminated.

There has been a lot of talk about the AFL expanding its final series to 10 teams, particularly after it moves to 19 teams when Tassie enters in a few years time, and then possibly 20. If so, I think the AFL could do a lot worse than just go with two Final Five conferences, with the winners playing off in the GF.

Nick Callil
October 17, 2024

Tony - of the factors you mention affecting the premiership probabilities in the footnote, the only one that could be quantified (albeit imperfectly) is home ground advantage (HGA). There's a website (https://www.aussportstipping.com/sports/afl/home_advantage/) which shows that on available records the median HGA for active teams is around 13%. So I've set out below the reworked the premiership probabilities allowing for these (with nil HGA for the grand final as it's played at that 'neutral' ground the MCG!) Far from perfect analysis but it suggests the the system is a little fairer to higher ranked teams than is sometimes claimed.

Home Ground Advantage
Team Nil 5% 10% 13%
1 18.75% 20.69% 22.80% 24.16%
2 18.75% 20.69% 22.80% 24.16%
3 18.75% 19.18% 19.20% 19.00%
4 18.75% 19.18% 19.20% 19.00%
5 6.25% 5.57% 4.80% 4.31%
6 6.25% 5.57% 4.80% 4.31%
7 6.25% 4.56% 3.20% 2.53%
8 6.25% 4.56% 3.20% 2.53%

Tony Dillon
October 17, 2024

Interesting Nick. And looking at the premiership winners in the 25 years to 2024 that the current final 8 system has been in-force, yields the following distribution:

Team Wins %age?
1 9 36.00% ?
2 7 28.00%?
3 7 28.00%?
4 0 0.00%?
5 1 4.00% ?
6 0 0.00% ?
7 1 4.00% ?
8 0 0.00%

A small sample size, but this distribution is closer to your HGA 13% scenario than the others, and implies winning probabilities greater than 50% in the majority of top 4 sides’ finals. No doubt because they are top 4 sides, and probably also because there are home ground advantages. Curiously, a team finishing 4th has never won it. Watch out for that in future. A poisoned chalice!

John
October 11, 2024

A terrific analysis and something I have never seen before

John
October 11, 2024

Be interesting to see for over the years what the rise of AFL in Australia would be compared to NRL. On a national basis. Using similar sorts of calcs and metrics. Based on things such as viewership via the various media types, the press and attendances. There may be other metrics I haven't thought of.

Sean
October 10, 2024

Interesting read outcomes here. Thanks for including some data and statistics here that aren't investment related. Roll on the next footy code season - the A-League!

Neil
October 10, 2024

Causation and correlation are not the same thing. What other factors could have CAUSED the outcome? If you ran the same analysis on the NRL (with no last round bye for all teams), what would the results be?

Martin Mulcare
October 14, 2024

Fair challenge, Neil, and I have looked up the corresponding NRL experience. Note that the current finals system has only been in place since 2012. For the most recent 13 seasons:
1. The proportion of teams that won their first final and then exited in the preliminary final was 23.1% (6 occasions out of 26), in between Tony's AFL ratios of 12.5% (no bye) and 43.75% (with a last week bye).
2. The proportion of "straight set exits" was 10.5% (6 incidences out of 52), also in between Tony's AFL ratios of 7.8% (no bye) and 21.9% (with a last week bye).
And for older rugby league fans, there was similar concern about the minor premiers potentially having two weeks off under the "Final Five" system that ran from 1973 -1994. My quick study suggests that the minor premiers had more problems with the first week off (12 out of 22 lost the "major semi final") than the second week off (7 out of the 10 minor premiers that won their major semi went on to win the grand final.

 

Leave a Comment:


banner

Most viewed in recent weeks

Vale Graham Hand

It’s with heavy hearts that we announce Firstlinks’ co-founder and former Managing Editor, Graham Hand, has died aged 66. Graham was a legendary figure in the finance industry and here are three tributes to him.

The nuts and bolts of family trusts

There are well over 800,000 family trusts in Australia, controlling more than $3 trillion of assets. Here's a guide on whether a family trust may have a place in your individual investment strategy.

Welcome to Firstlinks Edition 583 with weekend update

Investing guru Howard Marks says he had two epiphanies while visiting Australia recently: the two major asset classes aren’t what you think they are, and one key decision matters above all else when building portfolios.

  • 24 October 2024

Warren Buffett is preparing for a bear market. Should you?

Berkshire Hathaway’s third quarter earnings update reveals Buffett is selling stocks and building record cash reserves. Here’s a look at his track record in calling market tops and whether you should follow his lead and dial down risk.

Preserving wealth through generations is hard

How have so many wealthy families through history managed to squander their fortunes? This looks at the lessons from these families and offers several solutions to making and keeping money over the long-term.

A big win for bank customers against scammers

A recent ruling from The Australian Financial Complaints Authority may herald a new era for financial scams. For the first time, a bank is being forced to reimburse a customer for the amount they were scammed.

Latest Updates

Shares

Looking beyond banks for dividend income

The Big Four banks have had an extraordinary run and it’s left income investors with a conundrum: to stick with them even though they now offer relatively low dividend yields and limited growth prospects or to look elsewhere.

Exchange traded products

AFIC on its record discount, passive investing and pricey stocks

A triple headwind has seen Australia's biggest LIC swing to a 10% discount and scuppered its relative performance. Management was bullish in an interview with Firstlinks, but is the discount ever likely to close?

Superannuation

Hidden fees are a super problem

Most Australians don’t realise they are being charged up to six different types of fees on their superannuation. These fees can be opaque and hard to compare across different funds and investment options.

Shares

ASX large cap outlook for 2025

Economic growth in Australia looks to have bottomed, which means it makes sense to selectively add to cyclical exposures on the ASX in addition to key thematics like decarbonisation and technological change.

Property

Taking advantage of the property cycle

Understanding the property cycle can be a useful tool to make informed decisions and stay focused on long-term goals. This looks at where we are in the commercial property cycle and the potential opportunities for investors.

Investment strategies

Is this bedrock of financial theory a mirage?

The concept of an 'equity risk premium' has driven asset allocation decisions for decades. A revamped study suggests it was a relatively short-lived phenomenon rather than the mainstay many thought.

Vale Graham Hand

It’s with heavy hearts that we announce Firstlinks’ co-founder and former Managing Editor, Graham Hand, has died aged 66. Graham was a legendary figure in the finance industry and here are three tributes to him.

Sponsors

Alliances

© 2024 Morningstar, Inc. All rights reserved.

Disclaimer
The data, research and opinions provided here are for information purposes; are not an offer to buy or sell a security; and are not warranted to be correct, complete or accurate. Morningstar, its affiliates, and third-party content providers are not responsible for any investment decisions, damages or losses resulting from, or related to, the data and analyses or their use. To the extent any content is general advice, it has been prepared for clients of Morningstar Australasia Pty Ltd (ABN: 95 090 665 544, AFSL: 240892), without reference to your financial objectives, situation or needs. For more information refer to our Financial Services Guide. You should consider the advice in light of these matters and if applicable, the relevant Product Disclosure Statement before making any decision to invest. Past performance does not necessarily indicate a financial product’s future performance. To obtain advice tailored to your situation, contact a professional financial adviser. Articles are current as at date of publication.
This website contains information and opinions provided by third parties. Inclusion of this information does not necessarily represent Morningstar’s positions, strategies or opinions and should not be considered an endorsement by Morningstar.