Brownlow Medal 2020 Predictions

Background

First awarded to Edward Greeves from Geelong in 1924, the Brownlow Medal is the highest individual honour that can be bestowed upon an AFL footballer. Based on performance in each of the home and away matches played for the season, votes are assigned to the three best players (3 – first, 2 – second, 1- third) by the umpires that preside over the game. With suspended players deemed ineligible to win the medal, the Brownlow is awarded to the player perceived by umpires to be both the ‘best’ and ‘fairest’ for the season.

2020 Brownlow Medal

The 2020 AFL football season has been a year like no other. For the past 50 years, players have had 22 games in which to poll Brownlow votes. In 2020, a global pandemic has cut this number of games to 17. Not since the Second World War (1943) has a Brownlow medal been decided with players having fewer opportunities to vote.

In his 7 seasons at Fremantle (2012-2018), Lachie Neale polled 63 Brownlow votes from 128 games at an average of just under 0.5 votes per game. Since moving to Brisbane he has raised his game to a new level finished equal 3^rd in the 2019 Brownlow with 26 votes at an average of 1.18 votes per game include votes in 8 of the first 9 games of the year. In 2020 he has continued to improve and is easily the player most likely to win the 2020 Brownlow medal. Neale is ranked as the leading player on the ground on 8 occasions and is ranked in the top 3 a further 3 times.

Prior to 2020, Travis Boak had polled 118 Brownlow votes from 255 matches, placing him in the top 10 list of most votes by current players. With 4 top 20 finishes in the Brownlow (including 2 top 10 finishes) he has been performed at an elite level for more than a decade. With Port Adelaide finishing the season atop the AFL ladder and Boak ranked as the leading player on the ground on 6 occasions, he is likely to finish in the top 3.

In stark contrast to Boak, St Kilda midfielder Jack Steele is a relative Brownlow ‘novice’ having polled only 5 votes (from 78 games) prior to 2020.  At 25 years of age, and with 97 matches now under his belt, Steele is approaching his footballing prime. In 2020 he has played 6 of the best 7 ranked games of his career and has a chance of polling votes in 9 matches. Assuming the umpires view his performances as highly as the coaches, he is a good chance for a podium finish.

Melbourne onballer Christian Petracca has a similar Brownlow profile to Steele, having also only ever polled 5 career votes. Prior to 2020, Petracca had started less than 25% of matches as a midfielder (21/85 games). A strong pre-season, however, has seen him start more than 70% of matches in 2020 (12/17 games) as an onballer and his ability to influence games has increased accordingly with 9 out of 10 career-best games coming this season. He is ranked in the top 3 players on the ground 7 times and despite having never polled 3 votes in the past, is heavily fancied by football experts and bookmakers alike to finish in the top 3.

Western Bulldog midfielders Jack Macrae and Marcus Bontempelli have both polled votes in the same match on 14 occasions with Macrae polling more votes in only 2 of these matches. Despite playing in the shadow of Bontempelli, Macrae is an elite Brownlow performer in his own right with 66 votes from 135 games at an average just under a half a vote, placing him in the top 20 list of current players for votes per game. In 2019 he polled 22 votes from 22 games, finishing equal 8^th with Bontempelli. With 8 games ranked in the top 3 in 2020, expect Macrae to outvote Bontempelli in 2020.

Following what was labelled “a fairly innocuous incident”, Essendon midfielder Zach Merrett was suspending for 1 match following the round 4 clash with Carlton and is thus ineligible to win to 2020 Brownlow medal. However, with 5 potential best on ground performance in 2020, he is likely to poll well.

Expect the remaining positions in the top 10 to be filled by perennial Brownlow vote catchers Dangerfield, Martin, Bontempelli and Parker.

Leading Players for 2020

Ranked 1^st  =  Number of times each player was ranked as the leading player on the ground

Ranked Top 3 = Number of times each player was ranked in the top 3 players on the ground

Ranked Top 5 = Number of times each player was ranked in the top 5 players on the ground

*ineligible

Predicting the Brownlow Medal Winner

By using past information, it is possible to identify a specific match and player features that are related to who will be awarded votes. For example, over the previous 5 years, 90% of all 3-votes, 77% of all 2-votes and 72% of all 1-votes have been awarded to players from the winning team, with the probabilities further increasing with the margin of victory. Similarly, the number of possessions that each player gets is strongly related to the number of votes polled, with the leading possession winner for each match receiving votes in 2 out of every 3 games over the past 5 years.

The natural ordering of the voting structure (3-2-1-0) lends itself to the use of a mathematical modelling process known as ordinal logistic regression. With a wealth of player information now readily available, by using information from over 2000 past matches, it is possible to identify and combine the effects of more than 20 contributing variables to produce a probability that each player will poll 3, 2 or 1 vote in each match.

The main contributing variables in the model include features of the game such as kicks, marks, handballs, goals and margin of victory, however, features relating to the specific player such as quality of possession and past voting performance are also important contributors.  By combining these probabilities, each player is assigned a score from 0 to 3 for each match. By aggregating these scores over the entire season, it is then possible to assign each player with a predicted total of Brownlow votes for the season and with multiple simulations, a probability of success can then be obtained.

How well do the models perform?

Per-game Level

The model is remarkably good at identifying the leading players in each match, with the top 5 ranked players accounting for 97% of all 3-votes, 90% of all 2-votes and 77% of all 1-votes over the past 5 years. Furthermore, the leading ranked player in each match as determined by the modelling process has successfully poll 3 votes 65% of the time, 2 votes 19% of the time and 1 vote 9% of the time. When matches are further categorised according to whether there was 1 clear standout player (as opposed to 2 or 3 standout players) the leading ranked player has successful poll 3 votes in 74% of all games.

Per-season Level

When individual match scores are aggregated for the season, the modelling process can accurately predict the season tally of votes for 84% of all players to within 1 vote. However, it is important to note that more than 70% of all footballers each season will not poll any votes at all. When considering the top 10 leading players each season, the modelling process will typically identify 7 or 8 of the top 10 players and 3 or 4 of the top 5. When considering the winner, over the past 17 seasons, all winners have come from the top 4 ranked players each year, with the leading ranked player winning on 9 occasions, the second-ranked player winning on 5 occasions and the third-ranked player winning twice.

How The Count Will Pan Out On The Night

Expect the favourite to start well with Neale likely to poll votes in the first 5 games of the season.

While Travis Boak may stick with him for the first few rounds it is unlikely that Neale will be headed from Round 5 onwards and should enjoy a 2-3 vote buffer for the duration of the count.

How the count will progress for leading players in the 2020 Brownlow medal

Actual and predicted order and vote totals for the top 10 players for the past seven years

By Team

Historically, when considering individual teams, the leading vote-getter for each team was ranked first by the model 85% of the time, with the 2^nd ranked player by the model getting the most votes 11% of the time.

Leading Players for Each Club for the 2020 Season

Ranked 1^st  =  Number of times each player was ranked as the leading player on the ground

Ranked Top 3 = Number of times each player was ranked in the top 3 players on the ground

Ranked Top 5 = Number of times each player was ranked in the top 5 players on the ground

Team Totals

By aggregating all individual player scores to a team level it is possible to derive a predicted vote tally for each team. The team to poll the most votes usually aligns with the team to win the most matches and this is the case in 2020 with Port Adelaide the team most likely to receive the most votes.

Team Polling the Most Votes in 2020