ATAR Estimator

ATAR Estimator helps students understand the likely range and scaling direction of their current subject results. It is available at /app/atar-estimator on Scholar plans that include the RoxWhy study-tools entitlement.

The estimator is not an official ATAR calculator. Official ATARs depend on the graduating cohort, final subject distributions, authority rules, and yearly scaling data. RoxWhy therefore shows a range, confidence, data freshness, model version, source references, and caveats instead of a single guaranteed number.

What it supports

• QLD, NSW, ACT, and VIC.
• Single score or score-range entry.
• State-specific subject selectors.
• Optional import from Academic Tracker to prefill matching subjects and current percentages.
• Eligibility feedback, used-subject breakdown, aggregate range, ATAR range, scaling direction, and confidence.
• Advanced forecasting with input source, assessment completion, confidence, likelihood bands, and optional saved forecasts.

How estimates work

QLD

RoxWhy validates the entered subjects and sends only QTAC subject codes and score ranges through a server adapter to the public QTAC Navigator estimate service. It never sends user ID, email, school, tracker IDs, or other personal account data. If QTAC is unavailable, RoxWhy shows an unavailable result with a link to QTAC Navigator rather than inventing a number.

NSW

RoxWhy maps entered HSC marks or mark ranges to approximate scaled-mark ranges using published UAC-style scaling summaries, then applies the best two English units plus the best eight remaining eligible units. It maps the aggregate range against recent aggregate-to-ATAR reference points.

VIC

RoxWhy interpolates VCE study scores through VTAC-style scaling points, applies English plus the next best three studies, and adds eligible 10 percent increments. It then maps the resulting aggregate range to a recent VTAC-style aggregate-to-ATAR table.

ACT

ACT raw school percentages cannot be transformed accurately without college, course, and AST scaling context. If only raw percentages are entered, RoxWhy asks for scaled course scores. When scaled course scores are supplied, it applies the ACT-style aggregate rule of the best three major scaled course scores plus 0.6 of the next best course.

Data freshness

RoxWhy stores refreshable ATAR model payloads for local NSW, VIC, and ACT estimates. Super admins can refresh the public source snapshots from the ATAR data operations page. The refresh process stores source URLs, report years, fetch timestamps, checksums, discovered report links, model versions, and model metrics so students can see when the model was last updated.

Model refreshes create candidate models for review. A super admin can run a backtest and then activate a reviewed model. This keeps newly fetched data from reaching students before source coverage, caveats, and model metrics have been checked.

Some source tables are seeded from published report-style points while the admin refresh pipeline gathers source snapshots. Those rows are labelled with caveats and reliability so students know when a value is modelled rather than directly official.

Advanced forecasting and consent

Advanced mode asks whether each input is a current course percentage, trial mark, expected final mark, or official scaled score. RoxWhy then runs a deterministic forecast simulation and reports likelihood bands such as P10, P50, and P90 rather than a single exact number.

Students can choose to save a forecast. Saved forecasts are private to the account and can later be matched with a final outcome submission. Model-improvement data is opt-in: RoxWhy only uses de-identified aggregate signals after the student has opted in, and user-derived cohort adjustments are not applied unless a jurisdiction, subject, and score-band group reaches the privacy threshold.

Why results are ranges

ATAR scaling changes with each cohort. A student can enter useful current marks and still not know the final cohort distribution, the final school/course moderation context, or all authority adjustments. RoxWhy therefore widens estimates when the source data is less direct, the entered score is a range, or the state requires context that the student has not supplied.

Use the result to identify likely scaling risk, compare subject contribution ranges, and decide which weak subject to plan around next. Do not treat it as an official admissions result.