BECOME AN AFL ANALYST

Meet James Harrold, learn how the course runs, and see exactly what you'll have built by the end of Week 6.

• Meet your instructor
• Course structure
• Assignments and course project overview

Map the AFL data landscape across event, GPS, and tracking sources, and see how analytics has shifted coaching from the eye test to data-informed decision-making.

• AFL data landscape and data source types
• Three key data types
• How coaches and football departments use analytics weekly
• Case study: Moneyball's influence on the AFL, with an interactive data-volume poll
• Data's role in the AFL ecosystem

Take a raw, messy post-match data export and turn it into a clean, analysis-ready dataset in R, comparing public fitzRoy data against Champion Data.

• R refresher: filtering, grouping, summarising with Tidyverse
• Public fitzRoy data vs Champion Data structures
• Demo: Live-coding session cleaning a messy post-match data dump
• Workshop: Structuring Centre Bounce Attendances (CBAs) from provided datasets

Assignment #1: CBA Summary Table
Clean a second mock match export into a tidy dataset and produce a CBA summary table, opening with a five-line data profile of the structure.

Break a match into possession chains and use their origins, locations, and outcomes to answer the core tactical questions coaches ask about how the ball moves.

• Champion Data metric terminology and definitions
• Corridor vs boundary usage and Inside 50 efficiency
• Score sources
• Stoppage and clearance profiling for team benchmarks
• Demo: Quadrant-style game-play analysis, fast vs slow, contest vs uncontested
• Workshop: Mapping real possession sequences that lead to a score

Build an evidence-based opposition profile, identifying an opponent's strengths and weaknesses from data and turning them into a scouting report a coaching group can act on.

• Traditional vs analytics-based scouting
• Structural anomalies and what makes a player "good" at their position
• Match-up optimisation for key position battles
• Workshop: Drafting a pre-game scouting report on a provided opponent

Understand the key physical performance metrics and devices used in AFL, and identify which metrics actually contribute to team and player success.

• Demo: Raw Catapult/STATSports export to a player workload summary
• Total distance, High-Speed Running, sprint efforts, and player load
• Speed thresholds and why velocity bands differ between clubs
• Match demands vs training intensity
• Case study: Positional GPS patterns in rugby union backs vs forwards, tested against AFL midfields
• Workshop: Calculating High-Speed Running and total distance from a clean GPS file

Assignment #2: Positional Workload Comparison
From a multi-player GPS file, produce a positional workload comparison and three observations written for a high-performance manager.

Use data to compare and project players, translating Under-18 and state league numbers to AFL level, and build visual comparisons that inform draft and trade decisions.

• Translating Talent League and state league data to AFL level
• Player similarity models for draft forecasting and list demographics
• Case study: Analysing a recent AFL blockbuster trade with historical data
• Workshop: Building radar charts to evaluate draft prospects against AFL players

Understand how predictive models are built, construct a team rating system and win-probability model from real match data, and use the outputs to evaluate teams and players.

• What a model is and how it predicts sporting outcomes
• How Elo ratings work, and converting an Elo gap into win probability
• Performance benchmarks above replacement level
• Stability and reliability of AFL metrics year to year
• Workshop: Completing an Elo update loop to rate all 18 teams and predict the round

Assignment #3: Elo Ratings Table
Submit an 18-team Elo ratings table and predictions with win percentages for the current round.

Apply data-visualisation principles to turn your analysis into a simple, coach-ready dashboard built in R Shiny.

• Data-visualisation best practices for sport: colour, clutter, cognitive load
• Introduction to R Shiny for football data
• Demo: Live critique of real sports dashboards, good vs bad
• Automating recurring reporting
• Demo: Connecting a clean dataset to a dynamic team and player profile page
• Workshop: Building ggplot visuals with labelling and colour coding

Assignment #4: Wireframe Dashboard
Prepare a wireframe dashboard ready for deployment, covering information from every section of the high-performance department.

Understand what x,y coordinates of players and the ball unlock, and visualise movement and space from real tracking coordinates.

• Optical tracking data and XY coordinates
• Field control and space-creation models
• Measuring defensive zone integrity
• Limitations of current broadcast data and future frontiers
• Workshop: Converting lat/long to pitch x,y with the Haversine formula and plotting a play frame by frame

Turn analysis into a message a coach will actually listen to: keep it short, clear, and ready to handle the questions that come back.

• Coach and player psychology, and working within their time constraints
• Managing resistance to analytics in traditional football environments
• Presenting the right visual to the right audience
• Workshop: Presenting complex or longitudinal insights to non-technical staff in under two minutes

Leave with a public sports analytics portfolio taking shape and a clear map of the industry job market, from structuring a GitHub repository to handling technical interviews.

• Structuring a GitHub repository or public portfolio
• Navigating the sports industry job market and technical interviews
• Workshop: Peer review and group feedback on Course Project dashboard drafts

Selected students present their dashboards to the class, closing out the course with live evaluation and feedback.

• Data vs art: going beyond analytics in a club environment
• Instructor-selected Course Project presentations and live evaluations
• Course wrap-up