Brownlow Medal 2019 Predictions

About The Author

Professor Michael Bailey has worked as a biostatistician for 20 years and has published over 500 manuscripts in peer reviewed medical journals that have been cited more than 26,000 times. In 2005 he completed a PhD which outlined a statistical approach to predicting sporting outcomes and with a specific chapter dedicated to the Brownlow, he is the only known person to be awarded a PhD for the prediction of the Brownlow medal winner. He is back again this year with his 2019 Brownlow Medal 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.


2019 Brownlow Medal

Throughout the history of the Brownlow medal, there have been 13 multiple winners. Expect this number to grow in 2019 with past winners Nat Fyfe and Patrick Dangerfield the two players most likely to take home the medal. Using the historical ratio of votes polled per game played as a marker of quality, Nathan Fyfe and Patrick Dangerfield are currently the best two players in the competition. Averaging almost one vote per game (vpg), 2015 winner Nathan Fyfe has the highest average votes per game [0.99vpg (141 votes/142 games)] amongst current players and is second only to triple Brownlow medallist Hayden Bunton [1.04v/g (122/117)] on the all-time list.

While neither Fyfe nor Dangerfield has had the best year of their careers, both have played enough high ranking games to win a second Brownlow medal with each having a one in three chance of success. The fact that more than three quarters of all AFL footballers will never be awarded a 3-vote from umpires, indicates that it takes a certain class of footballer to be consistently recognised as the best player on the ground. Fyfe and Dangerfield share a trait that is common amongst Brownlow medallists, but few other players possess: an ability to poll more 3-votes than 2s and 1s combined (3-2-1 Ratio; Fyfe:32-17-11, Dangerfield:41-24-11). If there is to be a changing of the guard and one of these 2 past champions is to be usurped, then it is most likely to be by a player who also shares this trait.

In his seven seasons at Fremantle, Lachy Neale polled more 3s than 2s or 1s combined (3-2-1 ratio 15-5-8). Despite a proven ability to catch the umpire’s eye, Neale’s ability to poll Brownlow votes had been hampered by playing in the same team as Nat Fyfe (matches without Fyfe 0.59vpg (26/44) vs matches with Fyfe 0.44vpg (37/84)). Accordingly, his move to Brisbane has seen his chance of winning a Brownlow increase significantly and in 2019 he has a 15% chance of success in 2019.

Since being awarded best-on-ground in only his ninth game of AFL football (Round 15 2014), Marcus Bontempelli has carried the moniker “Future Brownlow medallist”. With a remarkable 3-2-1 ratio of 17-5-4, when Bontempelli plays well, he is the player the umpires will award the 3 votes to. With a good start to the year and a great finish, Bontempelli has a 10% chance of collecting his first Brownlow Medal, with a barren patch between rounds 8-14 the primary reason why his chance of success is not higher.

An off-the-ball elbow to Adam Kennedy in Round 3 will ensure that Dustin Martin will not win his second Brownlow medal in 2019. However, 7 outstanding games will be enough to ensure that he finishes high in the voting.

Jack Macrae is likely to be the best-performed player in the second half of the year and Tim Kelly is likely to be the best-performed player in the first half of the year. Whilst a lack of consistency will deny both players the medal, expect both to finish in the top 10.


Leading Players for 2019

Name Predicted Total Win % Place % Rank 1 Rank 3 Rank 5 Votes per game 3-2-1 ratio
Patrick Dangerfield 27 33 70 8 11 16 0.86 41-24-11
Nat Fyfe 27 33 70 6 15 16 0.99 32-17-11
Lachie Neale 24 15 40 7 10 15 0.49 15-5-8
Marcus Bontempelli 23 10 30 4 11 13 0.66 17-5-4
Dustin Martin 21 0 0 7 8 11 0.79 28-24-21
Jack Macrae 21 3 16 4 11 15 0.39 7-8-7
Tim Kelly 20 2 13 5 10 11 0.59 3-2-0
Patrick Cripps 20 2 13 5 9 13 0.60 7-11-6
Travis Boak 18 <1 8 4 9 12 0.44 16-21-12
Adam Treloar 18 <1 8 3 10 17 0.39 9-8-9


Predicting the Brownlow Medal Winner

By using past information, it is possible to identify specific match and player features that are related to who will be awarded votes. For example, over the previous five years, 89% of all 3-votes, 79% 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 possession that each player gets is strongly related to the number of votes polled, with the leading possession winner for each match receiving votes in two out of every three games over the past five 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 zero to three 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 five ranked players accounting for 97% of all 3-votes, 90% of all 2-votes and 77% of all 1-votes over the past five years. Furthermore, the leading ranked player in each match as determined by the modelling process has successfully poll 3 votes 63% of the time, 2 votes 20% of the time and 1 vote 9% of the time. When matches are further categorised according to whether there was one clear standout player (as opposed to two or three standout players) the leading ranked player will successful poll 3 votes in 72% 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 seven or eight of the top 10 players and three or four of the top five. When considering the winner, over the past 16 seasons all winners have come from the top four ranked players each year, with the leading ranked player winning on eight occasions, the second-ranked player winning on five occasions and the third-ranked player winning twice.


How The Count Will Pan Out On The Night

The 2019 Brownlow medal promises to be a close affair, with the winner unlikely to be determined until the final round of the season. Expect multiple leaders throughout the count with Dangerfield, Kelly, Bontempelli and Fyfe all leading at various stages. Dangerfield’s success will likely hinge on three potential best-on-ground performances in the last three rounds of the season.


How the count will progress for the leading 10 players


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

Year Player Finishing Order Predicted Order Actual Votes Predicted Votes
2018 T. Mitchell 1 1 28 29
S. Sidebottom 2 12 24 16
A. Brawshaw 3 22 21 12
M. Gawn 4 5 20 18
P. Cripps 5 8 20 17
D. Martin 6 2 19 25
R. Laird 7 9 19 17
D. Beams 8 23 18 12
J. Stevens 9 31 18 10
P. Dangerfield 10 3 17 20
2017 D. Martin 1 1 36 35
P. Dangerfield 2 2 33 34
T. Mitchell 3 3 25 29
J. Kennedy 4 5 23 20
L. Franklin 5 7 22 18
J. Kelly 6 6 21 20
R. Sloane 7 4 20 21
M. Bontempelli 8 8 19 17
O. Wines 9 12 18 15
D. Beams 10 26 17 12
2016 P. Dangerfield 1 1 35 35
L. Parker 2 12 26 17
D. Martin 3 2 25 23
R. Sloane 4 5 24 20
D. Hannebery 5 3 21 22
A. Gaff 6 7 21 18
A. Treloar 7 15 21 16
M. Bontempelli 8 8 20 18
L. Neale 9 19 20 15
R. Gray 10 9 19 18
2015 N. Fyfe 1 1 31 28
M. Priddis 2 3 28 24
S. Mitchell 3 2 26 27
J. Kennedy 4 4 25 22
D. Hannebury 5 5 24 22
P. Dangerfield 6 6 22 21
D. Martin 7 7 21 20
D. Mundy 8 11 19 16
C. Ward 9 26 19 11
T. Goldstein 10 9 18 20
2014 M. Priddis 1 4 26 20
N. Fyfe 2 1 25 25
G. Ablett 3 2 22 23
L. Franklin 4 12 22 15
J. Selwood 5 3 21 21
J. Kennedy 6 6 21 19
T. Boak 7 9 21 16
P. Dangerfield 8 17 21 14
S. Johnson 9 13 19 15
T. Cotchin 10 5 18 20
2013 G. Ablett 1 1 28 26
J. Selwood 2 3 27 24
D. Swan 3 2 26 25
S. Johnson 4 8 25 19
P. Dangerfield 5 11 22 17
S. Pendlebury 6 6 21 20
D. Hannebery 7 12 21 17
T. Rockliff 8 15 21 15
T. Cotchin 9 9 19 18
K. Jack 10 13 19 17


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 2nd ranked player by the model getting the most votes 11% of the time.

Leading Players for Each Club for the 2019 Season

Team Name Projected Total Ranked 1st Ranked Top 3 Ranked Top 5
AD M. Crouch 14 3 7 10
AD B. Crouch 11 1 6 10
BR L. Neale 24 6 9 15
BR D. Zorko 15 3 9 10
CA P. Cripps 20 5 9 13
CA M. Murphy 8 2 4 6
CO A. Treloar 18 2 7 16
CO B. Grundy 17 3 8 12
CO S. Pendlebury 16 3 9 12
ES Z. Merrett 17 2 8 9
ES D. Heppell 8 2 4 6
ES D. Shiel 8 1 3 8
WB M. Bontempelli 23 5 9 13
WB J. Macrae 21 3 9 16
FR N. Fyfe 27 6 14 14
FR M. Walters 13 4 6 8
GE P. Dangerfield 27 7 12 15
GE T. Kelly 20 5 8 11
GO B. Fiorini 5 1 3 4
GO D. Swallow 5 0 2 5
GW T. Taranto 14 3 6 11
GW J. Cameron 13 3 5 7
HA J. Worpel 11 2 5 6
HA J. Sicily 11 3 7 7
ME M. Gawn 14 3 8 13
ME C. Oliver 10 1 4 11
NO S. Higgins 14 2 6 9
NO B. Cunnington 12 1 8 10
PA T. Boak 18 4 8 11
PA R. Gray 10 2 4 8
RI D. Martin 21 7 8 11
RI D. Prestia 13 1 6 12
ST J. Billings 11 3 5 5
ST S. Ross 10 2 4 8
SY L. Parker 14 2 7 9
SY J. Kennedy 9 2 3 9
WE L. Shuey 15 4 8 10
WE A. Gaff 15 2 7 10
WE E. Yeo 14 1 9 11


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 correlates with the team to win the most matches and this is the case in 2019 with Geelong the team most likely to poll the most votes.

Team Polling the Most Votes in 2019

n Team Projected Total Win %
1 Geelong 91 72
2 Brisbane 83 9
3 Richmond 81 5
4 Collingwood 81 5
5 West Coast 81 4
6 GWS 80 4
7 Bulldogs 79 2
8 Port Adelaide 71 0
9 Fremantle 66 0
10 Hawthorn 66 0
11 Adelaide 63 0
12 North Melbourne 62 0
13 Essendon 59 0
14 Sydney 58 0
15 Carlton 52 0
16 St Kilda 50 0
17 Melbourne 41 0
18 Gold Coast 24 0


Related Articles

2019 Brownlow Medal Predictor

For your complete guide to the 2019 AFL Brownlow, check out Betfair's Predictor. You'll find assigned votes for every ...

Your 2019 Brownlow Medal Odds & Preview

Weekends dedicated to your partner are once again, over.  It’s time to realign those priorities in life to what’s ...

Your 2019 Brownlow Medal Hub

Our 2019 Brownlow Medal Minihub is your one-stop for all your Brownlow analysis and betting tips.