Matter of Stats

View Original

From One Season to the Next

With the 2024 Men’s AFL season just weeks away, I thought it timely to look at different perspectives of how teams have historically performed in home and away seasons from one season to the next.*

(* for different values of ‘next’)

PLAYING IN FINALS

Let’s firstly look at how many teams have made Finals in seasons separated by between 1 and 5 years. To allow like-for-like comparisons, we’ll only consider those seasons in which the top 8 teams played Finals, which locks us into the 2000 to 2023 time period.

The top row, for example, tells us that, of the 8 teams that played Finals in 2000, only 4 also played Finals in 2001. As well, 4 played FInals in 2002, 5 in seasons 2003 and 2004, and only 4 again played Finals in 2005.

One of the remarkable features of this table is that none of the numbers lie outside the range from 3 to 6, which implies that there is some commonality in the Finalists across seasons, but that it’s not complete.

The most recent example of 6 teams going back to back in Finals is 2019/2020 when Geelong, Brisbane Lions, Richmond, Collingwood, West Coast, and Western Bulldogs achieved the feat. The losing 2019 Grand Finalists, GWS, and Essendon, both missed the 2020 Finals being replaced by Port Adelaide and St Kilda.

On average, across the 23-season span, about 5.1 teams played Finals in successive years. Put another way, a team that plays Finals in one season is about a 64% chance of playing Finals again in the following season. Once we increase the gap to 2 years or more, the average number of repeating teams drops to about 4.7 (or around 59%)

If we restrict ourselves only to those years where 18 teams have competed, the equivalent averages are about 5.1 and 4.6. Were we to assume that teams’ ladder positions were determined randomly, the expected number of teams in common between the Finalists in any two seasons in an 18-team competition is 4/9 x 8 or about 3.6. As such, there is some positive predictive power in teams’ recent Finals history if we’re looking to determine which teams are most likely to play Finals in a given season - if a team played Finals, even as much, as 5 seasons ago, they have a better than random chance of playing Finals this season.

securing the double-chance

What if we restrict our analysis to teams finishing in one of the double-chance spots - that is, in 1st to 4th - across the same span of men’s AFL history?

Now we find a lot more volatilty, especially if we look more than a single season ahead.

In 2000, for example, we had Essendon, Carlton, Melbourne, and North Melbourne filling the top 4 spots. Of them, only Essendon finished in the Top 4 in 2001, and none of them finished in the Top 4 from 2002 to 2005.

(Remarkably, in fact, across the 20 seasons from 2001 to 2020, of those four teams only Essendon in 2001 and North Melbourne in 2007 ever finished in the Top 4 in a subsequent season)

There have been only two occasions when the same 4 teams filled the Top 4 spots in successive seasons, and the most recent of these was in 2013/2014 when Hawthorn, Geelong, Fremantle, and Sydney did it.

In contrast, there have been 4 occasions where none of the Top 4 from two seasons ago appeared in that season’s Top 4. The latest of these occurred in 2015 and 2017 The Top 4 for 2015 was Fremantle, West Coast, Hawthorn, and Sydney, who finished 14th, 8th, 12th, and 6th respectively in 2017 and were replaced at the top by Adelaide, Geelong, Richmond, and GWS.

Across the entire 23 seasons, on average 2.2 teams finished Top 4 in successive seasons (which equates to about a 59% probability), but this number declines quite rapidly as we increase the number of seasons we are looking ahead. At a 5-season gap, the average is down to 0.8 (or about 21%).

If we, once again, restrict ourselves only to those years where 18 teams have competed, the equivalent averages are about 2.4 and 0.9. Were we to assume that teams’ ladder positions were determined randomly, the expected number of teams in common between the Top 4 in any two seasons in an 18-team competition is 2/9 x 4 or about 0.9. As such, there is essentially zero predictive value in estimating a team’s Top 4 chances using whether or not they finished Top 4 exactly 5 seasons ago. There is, however, some positive predictive power in whether or not they finished Top 4 in any of the 4 preceding seasons.

RISES AND FALLS

For the final piece in this blog we’ll look not just at making or missing Top 8 or Top 4, but at the specific ladder positions in which teams finished in successive home and away seasons. We’ll also expand our view of history to take in every season where there has been 14 or more teams competing, which gives us 37 seasons (and 36 comparisons) from 1987 to 2023.

A few notes on this table (which might also help to explain it):

  • Minor Premiers in one season have also been Minor Premiers in the following season on 8 occasions, which is the highest for any ladder position, and have only failed to finish in the top 8 positions on the ladder on 4 occasions. The biggest fall was from 1st to 16th, which was recorded by Fremantle in 2015 and 2016, and the latest was Geelong in 2023 who finished 12th after being the 2022 Minor Premier. The average Minor Premier falls 3.4 spots in the following season.

  • Home and Away season Runners Up have elevated to Minor Premier in the following season on only 4 occasions, which is fewer times than for the teams from 3rd and 4th in the preceding season. Only 8 teams that finished 2nd in one season have not finished in the top 8 in the following season, the worst of them being Collingwood in 2003 and 2004, and Port Adelaide in 2007 and 2008, both of which fell to 13th. The average team from 2nd dropped 3.5 spots the following season.

  • Teams from 5th have recorded the largest average decline of 3.6 spots, having climbed the ladder on only 7 occasions, oddly never to 3rd and only once to 1st. Eight times they’ve finished 13th or worst, the nadir being Melbourne’s fall from 5th in 2018 to 17th in 2019.

  • Amongst teams finishing lower down on the ladder, it’s those from 14th who have recorded the largest average increase (3.7 spots), the best being Adelaide’s climb from 14th in 2011 to 2nd in 2012.

  • Teams from 16th have also fared relative well, climbing on average by 3.6 places. Best was Brisbane Lions’ 16th in 1998 to 3rd in 1999.

  • Since 2012 when we’ve had 18 teams, no team has finished 18th that finished higher than 11th in the preceding season.

  • Across the entire period, no team has finished as Minor Premier having ended the previous season’s home and away portion lower than 12th.

SUMMARY AND CONCLUSIONS

My high-level summary of all of this would be:

  • The League’s attempts to equalise the teams’ abilities have been fairly successful, though moreso in terms of changes in who are the very strongest teams (ie those finishing Top 4) than the teams who make Finals.

  • It’s been relatively rare for teams to record dramatic rises or falls in ladder position from one season to the next. About 40% of teams move by no more than two ladder positions up or down from one season to the next, and only 18% move by 7 or more ladder positions. The overall average absolute change in ladder position from one season to the next is just 3.8 ladder positions, and the most common movement is just a single ladder position, which has been the case for about 15% of all teams.