How we measure the quality of our injury records — and where the limits are.
Every injury report carries uncertainty — not even club medical teams know the full picture in real time. So instead of claiming we predict the future, we measure something honest: how well our records match what was publicly reported, how quickly they reflect new information, and how consistently we classify injuries against medical standards.
We publish confidence bands on uncertain reports ("TBC", "awaiting scans", "managed load") rather than pretending the ambiguity doesn't exist. Editorial judgement is human, not algorithmic.
We use four complementary checks:
How many of the injuries reported publicly each round end up in our database within 24 hours. We aim for 95%+ — measured by spot-checking AFL.com.au, club statements, and trusted reporters against our records.
Every injury we record is graded against the Orchard Sports Injury and Illness Classification System. We aim for 80%+ classified within a week — the rest stay flagged as "awaiting classification" until we have enough information.
Records are refreshed daily; status changes (active → fitness test → return) are reconciled across sources. A weekly editorial pass reviews new injuries, recoveries, and confidence bands by hand — automation does the heavy lifting, but a human signs off.
Where reports are vague ("TBC", "scans", "managed"), we tag the record with a confidence level rather than guessing a return date. Honest uncertainty beats false precision.
Every injury record carries a diagnosis phase — it enters as pending, hardens to provisional as reports land, and reaches confirmed once the OSIICS classification is settled (a small set close out as unclassifiable when public reporting never sharpens). Right now 30% of the 695 injuries tracked this season are confirmed:
Load-managed rests are excluded. These are live counts from the record, refreshed with the page.
We hold ourselves to these benchmarks:
For detailed data sources and the editorial process, see our Methodology page.
Quality isn't a one-time exercise — it's an ongoing commitment. Every week we review what we captured, what we missed, and where the records disagreed with later reports. The aim is a database that gets cleaner, more consistent, and more useful every season.
Honest reporting requires acknowledging what we don't know. We work exclusively with publicly available data — AFL.com.au injury lists, official game statistics, club statements, reporter confirmations, and historical records. We don't have access to confidential medical assessments or internal club reports.
That means some details will be imprecise. Reported severities can change. Recovery timelines slip. Some injuries are managed quietly by clubs until they can't be hidden. We log what we can verify, flag what we can't, and update records as new information arrives.
Trust comes from consistency, not claims. We'd rather be the most thorough, most honest record of AFL injuries on the internet than the loudest one — and that means publishing our quality targets, hitting them, and saying so plainly when we miss.