Model insights · AFLW Best and Fairest
How the model has called it
No bookmaker or outside model publishes AFLW Best and Fairest vote predictions, so there is no field to compare against. Every season below was predicted by a model that never saw that season's votes.
Feature weights and out-of-sample accuracy · Plackett-Luce · each season held out · 2017–2025
How accurate is the model
Top 10
100%
8/8
Top 5
100%
8/8
Winner in top 3
88%
7/8
avg vote error · top 10
3.1
Strong on the field, modest on the outright. The predicted top 10 is ranked tightly, but pinning the exact medallist past the podium is closer to a coin flip.
Season by season
| Season | Actual winner | Predicted rank | Avg error |
|---|---|---|---|
| 2025 | Ash Riddell | #1 | 4.2 |
| 2024 | Ebony Marinoff | #1 | 3.4 |
| 2023 | Monique Conti | #3 | 4.2 |
| 2022 S7 | Alexandra Anderson | #4 | 3.0 |
| 2022 S6 | Emily Bates | #1 | 2.2 |
| 2021 | Kiara Bowers & Brianna Davey | #1 | 2.8 |
| 2020 | Madison Prespakis | #2 | 3.7 |
| 2017 | Erin Phillips | #1 | 1.4 |
Predicted rank of the actual medallist. Emerald = top 3 · gold = outside top 10 (row tinted) · plain = 4–10. Each season predicted by a model fitted on the other seasons only, so none has seen its own votes.
Predicted vs actual votes · top 10 predicted
Gold = actual winner · emerald = other top-10 predicted · dashed line = perfect prediction
What drives Best and Fairest votes
goals_pct is the single strongest signal: 15.3% of total feature importance.
goals_pct15.3%
dreamTeamPoints_pct12.4%
metresGained_pct8.6%
scoreInvolvements_pct8.4%
coaches_pct8.1%
disposals_pct6.5%
goals_z5.5%
disposals_z4.6%
margin_c4.5%
tackles_pct4.2%
hitouts_pct2.5%
contestedPossessions_pct2.3%
win1.8%
inside50s_pct1.8%
contestedPossessions_z1.7%