The AFL is Australia’s top professional Australian-rules football competition, with 18 clubs across the country.
The sport mixes elements of football, rugby, and basketball-fast, physical, and heavily stats-driven, which makes it great for fantasy and data analysis.
Every season, analysts, coaches, and fans study millions of data points to understand why teams win, how players perform, and what tactics define success.
From GPS trackers in player jerseys to advanced AI models predicting match flow, data now drives every major AFL decision — from recruitment to strategy.
For fans and students of sport, this analytical shift has transformed the way we watch, measure, and interpret Australian rules football.
The Foundation: Key AFL Performance Metrics
At the heart of AFL analytics are its traditional statistics – the foundation for understanding player and team dynamics.
| Category | Description | Analytical Use |
|---|---|---|
| Disposals | Total of kicks + handballs | Measures involvement and ball movement. |
| Marks | Clean catches from a kick over 15m | Indicates control and spatial awareness. |
| Tackles | Defensive actions preventing possession | Tracks pressure and intensity. |
| Inside 50s | Entries into the attacking zone | Reflects offensive efficiency. |
| Clearances | Winning the ball after stoppages | Key for midfield dominance. |
| Hit-outs | Ruck taps after stoppages | Essential for possession control. |
| Contested Possessions | Winning the ball under physical pressure | Highlights strength and courage. |
| Efficiency % | Successful disposals ÷ total disposals | Evaluates decision-making and execution. |
Modern analysts use these metrics as inputs for broader models that estimate team strength, scoring potential, and match tempo.
Possession and Territory: The Core Data Framework
AFL matches are often described as a battle for territory and time. Data models simulate this through two key measurements:
Time in Possession
Teams that maintain the ball longer often control match tempo. Analysts study:
- Disposal chains (number of passes per possession).
- Turnover rates (how often teams lose the ball).
- Transition speed (time from defence to attack).
A high-possession side like Collingwood, for example, prioritises control and buildup, while fast-transition sides like Brisbane thrive on speed and chaos.
Territory Advantage
Another essential model measures where the ball spends most of the game.
- Inside-50 dominance indicates sustained pressure.
- Midfield intercepts suggest tactical control.
- Defensive rebound-50s show counterattack strength.
Together, possession and territory metrics predict which team dictates play – a crucial component in AFL tactical analysis.
Player Impact Models: Measuring Influence Beyond Stats
Not all possessions are equal. A 20-metre kick under pressure means more than an easy sideways handball.
This insight led to the development of Player Impact Models, such as the Champion Data Player Ratings system.
Champion Data’s Model (Used by the AFL)
Each player action is assigned a value based on:
- Field position
- Game context
- Pressure level
- Outcome (gain, loss, score involvement)
This creates an “expected impact” score for every player. A midfielder who wins a clearance under pressure rates higher than one who makes simple disposals in open play.
Educational takeaway: These models use expected value analysis – a concept borrowed from economics and probability theory – to quantify influence rather than volume.
Momentum Modelling: Predicting Match Flow
One of the AFL’s biggest innovations in sports science is momentum analysis — identifying when and why the energy of a match shifts.
Key Indicators Used:
- Consecutive scores within a timeframe.
- Time spent in attacking half.
- Defensive pressure acts and intercepts.
- Fatigue indicators (distance run, sprint frequency).
Data scientists model these as momentum curves, similar to stock market graphs, showing peaks and troughs in team dominance.
For coaches, these insights influence tactical substitutions and time management. For fans, they reveal how quickly a match can turn – especially late in quarters.
GPS and Biometric Data: Tracking Movement and Fatigue
Every AFL player wears a GPS tracker during training and matches. These devices record:
- Distance covered
- Sprint count and top speed
- Acceleration bursts
- Heat maps of field coverage
Analysts merge this data with heart rate monitors and load metrics to calculate player fatigue and work rate.
Example: If a midfielder’s sprint frequency drops by 20% in the second half, analysts might predict reduced clearance impact – vital information for coaching staff and physiologists.
In fantasy and educational analysis, this data explains why player performance fluctuates across quarters or fixtures.
Team Strength Models: Elo and Power Ratings
AFL analysts often use Elo-based models (adapted from chess) to evaluate relative team strength.
How Elo Works:
- Every team starts with a base rating.
- Points are gained or lost based on match outcomes and opponent strength.
- Home advantage, injuries, and margin of victory adjust calculations.
After each round, Elo ratings provide a statistically grounded estimate of each club’s current power level.
This allows fans and researchers to:
- Track momentum over a season.
- Identify underperforming teams.
- Predict competitive balance.
in simple terms: Elo quantifies “form” better than ladder positions alone.
Shot Quality and Scoring Efficiency
Beyond total points, modern analysis focuses on expected scoring — how efficiently teams convert opportunities.
Expected Score (xScore)
Every shot at goal is graded by:
- Distance from goal
- Angle
- Defensive pressure
- Player fatigue
The model predicts the probability of a goal (for instance, 0.65 from 30m out). Over time, comparing actual vs expected performance reveals finishing quality.
| Metric | Meaning |
|---|---|
| Above xScore | Team or player finishes efficiently (accurate under pressure). |
| Below xScore | Missed opportunities or poor shot selection. |
This kind of analysis helps identify elite forwards and tactical shot creation – insights often discussed in professional commentary.
Opposition Analysis and Match-Ups
Data also drives pre-game strategy. Analysts create match-up matrices — comparing player attributes across teams to identify key battles.
For example:
- A fast defender may be assigned to counter a high-disposal small forward.
- A tall ruck pairing may be tested against a more mobile duo.
These comparisons rely on databases tracking player height, speed, marking ability, and defensive pressure acts.
Fun fact: AFL clubs employ data scientists to simulate match scenarios hundreds of times before each fixture, tweaking line-ups for maximum efficiency.
Machine Learning in AFL Analytics
The AFL’s data explosion has led to the adoption of machine learning (ML) to uncover hidden trends.
Algorithms trained on historical data can:
- Predict ball movement patterns.
- Identify likely turnover zones.
- Model injury risks based on workload.
These models don’t predict match results for betting — they assist coaching, fitness, and recruitment by providing deeper context for decision-making.
Example Use Cases:
- AI video tagging automates player performance review.
- Neural networks identify undervalued draft prospects.
- Regression models highlight tactical inefficiencies (e.g., overusing the corridor).
Fan Analytics: How Supporters Use Data
Fans and fantasy-league participants use many of the same models on a smaller scale.
Common uses include:
- Tracking form trends for player selection.
- Comparing disposal efficiency between midfielders.
- Analysing team match-ups before major rounds.
Platforms such as AFL Fantasy and SuperCoach now integrate official data feeds from Champion Data, allowing fans to experience professional-grade analytics safely and responsibly.
From GPS tracking and machine learning to expected goals and player impact models, every insight helps explain the beauty and chaos of the game — without removing the emotion that makes it special.
