How Often Does The Best Team Win The Flag?

/

Finals series are a significant part of Australian sporting life. No professional team sport I know determines its ultimate victor - as does, say the English Premier League - on the basis of a first-past-the-post system. There's no doubt that a series of Finals adds interest, excitement and theatre (and revenue) to a season, but, in the case of VFL/AFL at least, how often does it result in the best team being awarded the Flag?

To answer this question I need a way to identify the "best" team in a season, and for this purpose I'll be using my all-time team MARS Rating System, which you might recall rewards a team's performance on the basis of its victory of margin relative to the margin that might have been expected given the quality of its opponent and whether or not it was playing at home.

Using this System produces a time series of Ratings for each team, recalculated at the end of every contest in which it participates, and generally tracking in a range from about 950 to 1,050. All teams start their first season with a Rating of 1,000. The chart below plots the time series for every team, treating each of the pairs of South Melbourne and Sydney, North Melbourne and the Kangaroos, Footscray and the Western Bulldogs, and the Brisbane Bears and Brisbane Lions as the same team (which matters mostly because teams carry 47% of their previous end-of-season Rating into the next).

Scrutinising the Ratings since 2000 - the period of time for which many of us will have first-hand experience of AFL and the relative strengths of the participating teams - I think broadly confirms their validity.

To return to the question at hand then, what do these Ratings tell us about the frequency with which the team Rated highest at the end of a season tends to collect the Flag in that season? It tells us, as reflected in the chart at right, that this occurs over two-thirds of the time.

That same chart reveals that teams Rated second-highest win about one-third as often and so, between them, the two highest-Rated teams account for almost 90% of the Flags across VFL/AFL history. Over three-quarters of the remaining 10% have been won by the third-highest Rated team, leaving just two to be collected by fourth-highest ranked teams (Adelaide in 1997 and Fitzroy in 1898) and just one by a fifth-highest ranked team (Melbourne in 1900).

The dominance of teams Rated first or second is greater still if the first 50 years or so of the competition is excluded. Since 1948 only two teams Rated lower than second on MARS Ratings have won the Premiership - Adelaide, as just noted, in 1997, and Sydney who were Rated third in 2005.

Runners-Up have come from a wider set of end-of-season MARS Rankings, but it's still the case that no team Ranked lower than eighth by MARS has finished as Runner-Up, and that teams from Ranks below fourth have only 11 such titles to their names. Just two of those 11 belong to teams that were Ranked seventh (Melbourne in 1946) or eighth (Melbourne in 1988).

It's reassuring to note that teams Ranked second have been the teams most likely to collect the Runners-Up medals in the GF. They've done so almost twice as often as highest-Rated teams and about 2.5 times as often as third- and fourth-highest Rated teams.

So, despite the randomness introduced by determining ultimate success on the basis of a very small set of games, teams that have proven their quality throughout the season - those Rated first or second by MARS - tend to do comparatively well nonetheless. Which is, I think, as we should want it to be.

Absent a Ratings System we might instead use the final home-and-away season ladder as a means of ranking teams within a season and then ask how often the Minor Premiers and Runners-Up in the minor premiership go on to compete in and win the Grand Final.

The charts below provide the answers in the same format as we used earlier and reveal that final ladder position is a poorer indicator of Finals success that are MARS Ratings (although, to be fair, I should note that the MARS Ratings are updated throughout the course of the Finals and the Rankings I've used above refer to post-Finals Ratings).

Only 54% of Minor Premiers have collected the Flag in the same year, and that includes the two Flags awarded in 1897 and 1924 to the Minor Premiers without the need for a Grand Final. Runners-up for the Minor Premiership have won the Flag about 29% of the time, which means that, combined, teams finishing first or second on the home-and-away season ladder have won 83% of Flags. That's about 7% less than the proportion of Flags won by the highest and second-highest MARS Rated teams.

This means, of course, more Flags for teams from lower on the ladder. Teams from third have won about 9% of Flags, from fourth about 6%, and from fifth and sixth less than 1% each. No team finishing lower than sixth on the ladder has ever won a Flag.

The same is true of Runners-up though the spread of titles across the teams from ladder positions 1 to 6 is quite different. Minor Premiers have finished as Runners-Up about 35% of the time while those teams from second on the ladder have finished Runners-up about 32% of the time.

Teams from fourth on the ladder have done slightly better than those from third, collecting the Runners-Up honours on 6 more occasions. Teams finishing fifth or sixth on the ladder have ended the season as Runners-Up twice each.

While the correlation between ladder position and Grand Final outcomes is weaker than that between MARS Ranking and Grand Final outcomes, final ladder positions and MARS Rankings are still quite strongly related, as evidenced by the heat map below which is based on a tabulation of these two dimensions.

About 40% of all teams have finished the season with the same MARS Ranking as final ladder position and almost 75% have finished with Rankings and ladder position differing by one place or less.

Only 20 teams in the history of the competition have finished a season with a home-and-away ladder ranking more than four places different from its MARS Ranking, some notable recent examples being:

  • Kangaroos (2013) - Ranked 5th by MARS but finishing 10th on the ladder
  • Port Adelaide (2008) - Ranked 8th by MARS but finishing 13th on the ladder 
  • Adelaide (2007) - Ranked 3rd by MARS but finishing 8th on the ladder 
  • Kangaroos (2007) - Ranked 10th by MARS but finishing 4th on the ladder 
  • Sydney (2007) - Ranked 2nd by MARS but finishing 7th on the ladder 
  • Kangaroos (2005) - Ranked 10th by MARS but finishing 5th on the ladder 
  • Melbourne (2005) - Ranked 12th by MARS but finishing 7th on the ladder 

To finish, here's a table of the season-by-season results showing the Premiers, Runners-Up and the highest Rated teams from every season. Note especially the success of the highest-Rated teams in recent times: they've won 14 of the last 20, and 30 of the last 37 Flags, including a 15-year streak from 1978 to 1992.